chore: release v2.78.0

Co-authored-by: Marc Klingen <git@marcklingen.com>
Co-authored-by: Max Deichmann <m.deichmann@tum.de>

.devcontainer/devcontainer.json

@@ -1,14 +0,0 @@
-{
-	"name": "langfuse dev",
-    "image": "mcr.microsoft.com/devcontainers/javascript-node:20-bookworm",
-    "features": {
-        "ghcr.io/devcontainers/features/docker-in-docker:2": {
-        "version": "latest",
-        "moby": true
-        }
-    },
-    "forwardPorts": [3000],
-
-    "onCreateCommand": "npm install -g pnpm@8.15.5",
-    "postCreateCommand": "cp .env.dev.example .env && pnpm run dx-f"
-}

.env.dev.example

@@ -6,6 +6,12 @@
 DIRECT_URL="postgresql://postgres:postgres@localhost:5432/postgres"
 DATABASE_URL="postgresql://postgres:postgres@localhost:5432/postgres"
 
+# Clickhouse
+CLICKHOUSE_MIGRATION_URL="clickhouse://localhost:9000"
+CLICKHOUSE_URL="http://localhost:8123"
+CLICKHOUSE_USER="clickhouse"
+CLICKHOUSE_PASSWORD="clickhouse"
+
 # Next Auth
 # You can generate a new secret on the command line with:
 # openssl rand -base64 32
@@ -44,4 +50,4 @@ REDIS_AUTH="myredissecret"
 
 LANGFUSE_WORKER_PASSWORD=mybasicauthsecret
 # openssl rand -hex 32 used only here
-ENCRYPTION_KEY=6c16874e5c0f0cc74ddec00425fa99fbe9ffbe412b7d5d906a4b00005df91403 
+ENCRYPTION_KEY=0000000000000000000000000000000000000000000000000000000000000000 

.env.local.example

@@ -18,4 +18,7 @@ REDIS_HOST="127.0.0.1"
 REDIS_PORT=6379
 REDIS_AUTH="myredissecret"
 
-LANGFUSE_WORKER_PASSWORD=myworkerpassword
+LANGFUSE_WORKER_PASSWORD=myworkerpassword
+
+# openssl rand -hex 32 used only here
+ENCRYPTION_KEY=0000000000000000000000000000000000000000000000000000000000000000

.env.prod.example

@@ -10,40 +10,47 @@ DATABASE_URL="postgresql://postgres:postgres@db:5432/postgres"
 # DIRECT_URL="postgresql://postgres:postgres@db:5432/postgres"
 # SHADOW_DATABASE_URL=
 # optional, set to true to disable automated database migrations on Docker start
-# LANGFUSE_AUTO_POSTGRES_MIGRATION_DISABLED= 
+# LANGFUSE_AUTO_POSTGRES_MIGRATION_DISABLED=
 
 # Next Auth
 # NEXTAUTH_URL does not need to be set when deploying on Vercel
 NEXTAUTH_URL="http://localhost:3000"
 
-# You can generate a new secret on the command line with:
+# For each of these, you can generate a new secret on the command line with:
 # openssl rand -base64 32
-# https://next-auth.js.org/configuration/options#secret
-NEXTAUTH_SECRET="secret"
-SALT="salt"
+NEXTAUTH_SECRET="secret" # https://next-auth.js.org/configuration/options#secret
+SALT="salt" # salt used to hash api keys
+
+# API level encryption for sensitive data
+# Must be 256 bits, 64 string characters in hex format, generate via: openssl rand -hex 32
+ENCRYPTION_KEY="0000000000000000000000000000000000000000000000000000000000000000"
 
 # Use CSP headers to enforce HTTPS, optional
-LANGFUSE_CSP_ENFORCE_HTTPS="true"
+# LANGFUSE_CSP_ENFORCE_HTTPS="true"
 
 
 # Docker only, optional
 # PORT=3000
 # HOSTNAME=localhost
 
-# Default project, optional
+# Default role for users who sign up, optional, can be org or org+project
+# LANGFUSE_DEFAULT_ORG_ID=
+# LANGFUSE_DEFAULT_ORG_ROLE=
 # LANGFUSE_DEFAULT_PROJECT_ID=
 # LANGFUSE_DEFAULT_PROJECT_ROLE=
 
+# Logging, optional
+# LANGFUSE_LOG_LEVEL=info
+# LANGFUSE_LOG_FORMAT=text
+
 # Enable experimental features, optional
 # LANGFUSE_ENABLE_EXPERIMENTAL_FEATURES=true
 
-# Disable automated database migrations on Docker start, optional
-# LANFUSE_WEB_MIGRATION_DISABLED=true
-
 # Auth, optional configuration
 # AUTH_DOMAINS_WITH_SSO_ENFORCEMENT=domain1.com,domain2.com
 # AUTH_DISABLE_USERNAME_PASSWORD=true
 # AUTH_DISABLE_SIGNUP=true
+# AUTH_SESSION_MAX_AGE=43200 # 30 days in minutes (default)
 
 # SSO, each group is optional
 # AUTH_GOOGLE_CLIENT_ID=
@@ -94,8 +101,31 @@ LANGFUSE_CSP_ENFORCE_HTTPS="true"
 # The page size can be adjusted if needed to optimize performance
 # DB_EXPORT_PAGE_SIZE=1000
 
-# Disable logging to `events` table in db
-# ENABLE_EVENT_LOG=false
+# Automated provisioning of default resources
+# LANGFUSE_INIT_ORG_ID=org-id
+# LANGFUSE_INIT_ORG_NAME=org-name
+# LANGFUSE_INIT_PROJECT_ID=project-id
+# LANGFUSE_INIT_PROJECT_NAME=project-name
+# LANGFUSE_INIT_PROJECT_PUBLIC_KEY=pk-1234567890
+# LANGFUSE_INIT_PROJECT_SECRET_KEY=sk-1234567890
+# LANGFUSE_INIT_USER_EMAIL=user@example.com
+# LANGFUSE_INIT_USER_NAME=User Name
+# LANGFUSE_INIT_USER_PASSWORD=password
+
+
+
+### START Enterprise Edition Configuration
+
+# Allowlisted users that can create new organizations, by default all users can create organizations
+# LANGFUSE_ALLOWED_ORGANIZATION_CREATORS=user1@langfuse.com,user2@langfuse.com
+
+# UI Customization Options
+# LANGFUSE_UI_API_HOST=https://api.example.com
+# LANGFUSE_UI_DOCUMENTATION_HREF=https://docs.example.com
+# LANGFUSE_UI_SUPPORT_HREF=https://support.example.com
+# LANGFUSE_UI_FEEDBACK_HREF=https://feedback.example.com
+
+### END Enterprise Edition Configuration
 
 
 
@@ -104,9 +134,6 @@ LANGFUSE_CSP_ENFORCE_HTTPS="true"
 # Disable the expensive analytics queries and related features
 # LANGFUSE_DISABLE_EXPENSIVE_POSTGRES_QUERIES="true"
 
-# Customize the default datetime offset filter used in for the UI tables
-# LANGFUSE_DEFAULT_TABLE_DATETIME_OFFSET="7" # int (days)
-
 ### END Envs to be deprecated in Langfuse v3.0
 
 
@@ -127,12 +154,12 @@ LANGFUSE_CSP_ENFORCE_HTTPS="true"
 # NEXT_PUBLIC_POSTHOG_KEY=
 
 # Sentry
+# NEXT_PUBLIC_LANGFUSE_TRACING_SAMPLE_RATE
 # NEXT_PUBLIC_SENTRY_DSN=
 # NEXT_SENTRY_ORG=
 # NEXT_SENTRY_PROJECT=
 # SENTRY_AUTH_TOKEN=
-# Used to determine the Sentry sample rate
-# LANGFUSE_TRACING_SAMPLE_RATE=
+# SENTRY_CSP_REPORT_URI=
 
 
 # Cloudflare Turnstile
@@ -140,16 +167,56 @@ LANGFUSE_CSP_ENFORCE_HTTPS="true"
 # TURNSTILE_SECRET_KEY=
 
 # Demo project that users can use to try the platform
+# NEXT_PUBLIC_DEMO_ORG_ID=
 # NEXT_PUBLIC_DEMO_PROJECT_ID=
 
 # Crisp chat
 # NEXT_PUBLIC_CRISP_WEBSITE_ID=
 
-# Encryption
-# ENCRYPTION_KEY=
-
 # Admin API
 # ADMIN_API_KEY=
 
+# Redis
+# REDIS_HOST=
+# REDIS_PORT=
+# REDIS_AUTH=
+# REDIS_CONNECTION_STRING=
+# REDIS_ENABLE_AUTO_PIPELINING=
+
+# Cache configuration
+# LANGFUSE_CACHE_API_KEY_ENABLED=
+# LANGFUSE_CACHE_API_KEY_TTL_SECONDS=
+# LANGFUSE_CACHE_PROMPT_ENABLED=
+# LANGFUSE_CACHE_PROMPT_TTL_SECONDS=
+
+# Rate limiting
+# LANGFUSE_RATE_LIMITS_ENABLED=
+
+# Stripe
+# STRIPE_SECRET_KEY=
+# STRIPE_WEBHOOK_SIGNING_SECRET=
 
 ### END Langfuse Cloud Config
+
+
+## START Langfuse V3 Ingestion 
+
+# Clickhouse
+# CLICKHOUSE_MIGRATION_URL=
+# CLICKHOUSE_URL=
+# CLICKHOUSE_USER=
+# CLICKHOUSE_PASSWORD=
+
+# Ingestion
+# LANGFUSE_INGESTION_BUFFER_TTL_SECONDS=
+# LANGFUSE_INGESTION_FLUSH_DELAY_MS= 
+# LANGFUSE_INGESTION_FLUSH_ATTEMPTS= 
+# LANGFUSE_INGESTION_FLUSH_PROCESSING_CONCURRENCY=
+# LANGFUSE_INGESTION_CLICKHOUSE_WRITE_BATCH_SIZE= 
+# LANGFUSE_INGESTION_CLICKHOUSE_WRITE_INTERVAL_MS= 
+# LANGFUSE_INGESTION_CLICKHOUSE_MAX_ATTEMPTS=
+# LANGFUSE_LEGACY_INGESTION_WORKER_CONCURRENCY=
+# LANGFUSE_ASYNC_INGESTION_PROCESSING="true"
+# QUEUE_CONSUMER_LEGACY_INGESTION_QUEUE_IS_ENABLED="true"
+
+## END Langfuse V3 Ingestion

.github/dependabot.yml

@@ -8,6 +8,7 @@ updates:
   - package-ecosystem: npm
     directory: "/" # Location of package manifests
     rebase-strategy: "disabled" # use dependabot-rebase-stale
+    open-pull-requests-limit: 10
     schedule:
       interval: "daily"
     versioning-strategy: "increase"
@@ -19,17 +20,25 @@ updates:
       - dependency-name: "@types/node"
       - dependency-name: "@trpc/*"
     groups:
-      sentry:
-        patterns:
-          - "@sentry/*"
       prisma:
         patterns:
           - "prisma"
-          - "@prisma/*"
+          - "@prisma/client"
       next:
         patterns:
           - "eslint-config-next"
           - "next"
-      patches:
-        update-types:
-          - "patch"
+      lodash:
+        patterns:
+          - "lodash"
+          - "@types/lodash"
+      observability:
+        patterns:
+          - "dd-trace"
+          - "@opentelemetry/*"
+          - "@appsignal/opentelemetry-instrumentation-bullmq"
+          - "@prisma/instrumentation"
+          - "@sentry/*"
+      radix-ui:
+        patterns:
+          - "@radix-ui/*"

.github/workflows/pipeline.yml

@@ -19,7 +19,7 @@ jobs:
       - uses: actions/checkout@v3
       - uses: pnpm/action-setup@v3
         with:
-          version: 8.15.5
+          version: 9.5.0
       - uses: actions/setup-node@v3
         with:
           node-version: 20
@@ -40,6 +40,9 @@ jobs:
       - name: Checkout
         uses: actions/checkout@v3
 
+      - name: Set NEXT_PUBLIC_BUILD_ID
+        run: echo "NEXT_PUBLIC_BUILD_ID=$(git rev-parse --short HEAD)" >> $GITHUB_ENV
+
       - name: Build and run both images from compose
         run: |
           docker compose -f docker-compose.build.yml up -d
@@ -63,6 +66,7 @@ jobs:
           fi
 
   tests-web:
+    timeout-minutes: 20
     runs-on: ubuntu-latest
     name: tests-web (node${{ matrix.node-version }}, pg${{ matrix.postgres-version }})
     strategy:
@@ -78,7 +82,7 @@ jobs:
       - uses: actions/checkout@v3
       - uses: pnpm/action-setup@v3
         with:
-          version: 8.15.5
+          version: 9.5.0
       - name: Use Node.js ${{ matrix.node-version }}
         uses: actions/setup-node@v3
         with:
@@ -93,7 +97,7 @@ jobs:
       - name: Load default env
         run: |
           cp .env.dev.example .env
-          cp .env.dev.example web/.env
+          grep -v -e '^REDIS_HOST=' .env.dev.example > .env
 
       - name: Run + migrate
         run: |
@@ -106,18 +110,32 @@ jobs:
       - name: Seed DB
         run: |
           pnpm run db:migrate
-          pnpm run db:seed
 
       - name: Build
         run: pnpm run build
+        env:
+          SENTRY_AUTH_TOKEN: ${{ secrets.SENTRY_AUTH_TOKEN }}
+          SENTRY_ORG: ${{ vars.SENTRY_ORG }}
+          SENTRY_PROJECT: ${{ vars.SENTRY_PROJECT }}
 
       - name: Start Langfuse
         run: (pnpm run start&)
+        env:
+          LANGFUSE_INIT_ORG_ID: "seed-org-id"
+          LANGFUSE_INIT_ORG_NAME: "Seed Org"
+          LANGFUSE_INIT_PROJECT_ID: "7a88fb47-b4e2-43b8-a06c-a5ce950dc53a"
+          LANGFUSE_INIT_PROJECT_NAME: "Seed Project"
+          LANGFUSE_INIT_PROJECT_PUBLIC_KEY: "pk-lf-1234567890"
+          LANGFUSE_INIT_PROJECT_SECRET_KEY: "sk-lf-1234567890"
+          LANGFUSE_INIT_USER_EMAIL: "demo@langfuse.com"
+          LANGFUSE_INIT_USER_NAME: "Demo User"
+          LANGFUSE_INIT_USER_PASSWORD: "password"
 
       - name: run tests
         run: pnpm --filter=web run test
 
   tests-worker:
+    timeout-minutes: 20
     runs-on: ubuntu-latest
     name: tests-worker (node${{ matrix.node-version }}, pg${{ matrix.postgres-version }})
     strategy:
@@ -133,7 +151,8 @@ jobs:
       - uses: actions/checkout@v3
       - uses: pnpm/action-setup@v3
         with:
-          version: 8.15.5
+          version: 9.5.0
+
       - name: Use Node.js ${{ matrix.node-version }}
         uses: actions/setup-node@v3
         with:
@@ -145,10 +164,17 @@ jobs:
         run: |
           pnpm install
 
+      - name: Install golang-migrate for Clickhouse migrations
+        run: |
+          curl -L https://github.com/golang-migrate/migrate/releases/download/v4.16.2/migrate.linux-amd64.tar.gz | tar xvz
+          sudo mv migrate /usr/bin/migrate
+          which migrate
+
       - name: Load default env
         run: |
           cp .env.dev.example .env
           cp .env.dev.example web/.env
+          cp .env.dev.example worker/.env
 
       - name: Run + migrate
         run: |
@@ -156,24 +182,33 @@ jobs:
           sleep 5 # Wait for PostgreSQL to accept connections
           docker compose ps
 
+      - name: Ensure no unhealthy status
+        run: |
+          if docker compose ps | grep "(unhealthy)"; then
+            echo "One or more services are unhealthy"
+            exit 1
+          else
+            echo "All services are healthy"
+          fi
+
       - name: Seed DB
         run: |
           pnpm run db:migrate
           pnpm run db:seed
+          pnpm run --filter=shared ch:up
 
       - name: Build
         run: pnpm --filter=worker... run build
 
       - name: run tests
         run: pnpm --filter=worker run test
-
   e2e-tests:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v3
       - uses: pnpm/action-setup@v3
         with:
-          version: 8.15.5
+          version: 9.5.0
       - uses: actions/setup-node@v3
         with:
           node-version: 20
@@ -209,10 +244,62 @@ jobs:
       - name: Run e2e tests
         run: pnpm --filter=web run test:e2e
 
+  e2e-server-tests:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v3
+      - uses: pnpm/action-setup@v3
+        with:
+          version: 9.5.0
+      - uses: actions/setup-node@v3
+        with:
+          node-version: 20
+          cache: "pnpm"
+          cache-dependency-path: "pnpm-lock.yaml"
+
+      - name: install dependencies
+        run: |
+          pnpm install
+
+      - name: Load default env
+        run: |
+          cp .env.dev.example .env
+          echo "LANGFUSE_ASYNC_INGESTION_PROCESSING=true" >> .env
+          echo "LANGFUSE_CACHE_API_KEY_ENABLED=true" >> .env
+          echo "LANGFUSE_CACHE_PROMPT_ENABLED=true" >> .env
+
+      - name: Run + migrate
+        run: |
+          docker compose -f docker-compose.dev.yml up -d
+          docker compose ps
+          sleep 5 # Wait for PostgreSQL to accept connections
+
+      - name: Seed DB
+        run: |
+          pnpm run db:migrate
+          pnpm run db:seed:examples
+
+      - name: Build
+        run: pnpm run build
+
+      - name: Run server
+        run: (pnpm run start&)
+
+      - name: Run e2e tests
+        run: pnpm --filter=web run test:e2e:server
+
   all-ci-passed:
     # This allows us to have a branch protection rule for tests and deploys with matrix
     runs-on: ubuntu-latest
-    needs: [lint, tests-web, tests-worker, e2e-tests, test-docker-build]
+    needs:
+      [
+        lint,
+        tests-web,
+        tests-worker,
+        e2e-tests,
+        test-docker-build,
+        e2e-server-tests,
+      ]
     if: always()
     steps:
       - name: Successful deploy
@@ -236,7 +323,7 @@ jobs:
     steps:
       - uses: pnpm/action-setup@v3
         with:
-          version: 8.15.5
+          version: 9.5.0
       - name: Setup node
         uses: actions/setup-node@v3
         with:
@@ -246,6 +333,9 @@ jobs:
       - name: Checkout
         uses: actions/checkout@v3
 
+      - name: Set NEXT_PUBLIC_BUILD_ID
+        run: echo "NEXT_PUBLIC_BUILD_ID=$(git rev-parse --short HEAD)" >> $GITHUB_ENV
+
       - name: Log in to the GitHub Container registry
         uses: docker/login-action@v2
         with:

.github/workflows/porter_stack_web_eu_temp.yml

@@ -1,29 +1,30 @@
 "on":
   push:
     branches:
-    - production
+      - production
 name: Deploy to web
 jobs:
   porter-deploy:
     runs-on: ubuntu-latest
     steps:
-    - name: Checkout code
-      uses: actions/checkout@v3
-    - name: Set Github tag
-      id: vars
-      run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
-    - name: Setup porter
-      uses: porter-dev/setup-porter@v0.1.0
-    - name: Deploy stack
-      timeout-minutes: 30
-      run: exec porter apply
-      env:
-        PORTER_CLUSTER: "4060"
-        PORTER_DEPLOYMENT_TARGET_ID: 3be075c0-b121-4123-aec1-5dc571626b0c
-        PORTER_HOST: https://dashboard.porter.run
-        PORTER_PR_NUMBER: ${{ github.event.number }}
-        PORTER_PROJECT: "12565"
-        PORTER_REPO_NAME: ${{ github.event.repository.name }}
-        PORTER_STACK_NAME: web
-        PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
-        PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4060 }}
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Set Github tag
+        id: vars
+        run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+      - name: Setup porter
+        uses: porter-dev/setup-porter@v0.1.0
+      - name: Deploy stack
+        timeout-minutes: 30
+        run: exec porter apply
+        env:
+          PORTER_CLUSTER: "4060"
+          PORTER_DEPLOYMENT_TARGET_ID: 3be075c0-b121-4123-aec1-5dc571626b0c
+          PORTER_HOST: https://dashboard.porter.run
+          PORTER_PR_NUMBER: ${{ github.event.number }}
+          PORTER_PROJECT: "12565"
+          PORTER_REPO_NAME: ${{ github.event.repository.name }}
+          PORTER_STACK_NAME: web
+          PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
+          PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4060 }}
+          PORTER_NEXT_PUBLIC_BUILD_ID: ${{ steps.vars.outputs.sha_short }}

.github/workflows/porter_stack_web_staging.yml

@@ -27,3 +27,4 @@ jobs:
           PORTER_STACK_NAME: web
           PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
           PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4037 }}
+          PORTER_NEXT_PUBLIC_BUILD_ID: ${{ steps.vars.outputs.sha_short }}

.github/workflows/porter_stack_web_us_temp.yml

@@ -27,3 +27,4 @@ jobs:
           PORTER_STACK_NAME: web
           PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
           PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4054 }}
+          PORTER_NEXT_PUBLIC_BUILD_ID: ${{ steps.vars.outputs.sha_short }}

.github/workflows/porter_stack_worker_eu_temp.yml

@@ -1,29 +1,29 @@
 "on":
   push:
     branches:
-    - production
+      - production
 name: Deploy to worker (EU temp)
 jobs:
   porter-deploy:
     runs-on: ubuntu-latest
     steps:
-    - name: Checkout code
-      uses: actions/checkout@v3
-    - name: Set Github tag
-      id: vars
-      run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
-    - name: Setup porter
-      uses: porter-dev/setup-porter@v0.1.0
-    - name: Deploy stack
-      timeout-minutes: 30
-      run: exec porter apply
-      env:
-        PORTER_CLUSTER: "4060"
-        PORTER_DEPLOYMENT_TARGET_ID: 3be075c0-b121-4123-aec1-5dc571626b0c
-        PORTER_HOST: https://dashboard.porter.run
-        PORTER_PR_NUMBER: ${{ github.event.number }}
-        PORTER_PROJECT: "12565"
-        PORTER_REPO_NAME: ${{ github.event.repository.name }}
-        PORTER_STACK_NAME: worker
-        PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
-        PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4060 }}
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Set Github tag
+        id: vars
+        run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+      - name: Setup porter
+        uses: porter-dev/setup-porter@v0.1.0
+      - name: Deploy stack
+        timeout-minutes: 30
+        run: exec porter apply
+        env:
+          PORTER_CLUSTER: "4060"
+          PORTER_DEPLOYMENT_TARGET_ID: 3be075c0-b121-4123-aec1-5dc571626b0c
+          PORTER_HOST: https://dashboard.porter.run
+          PORTER_PR_NUMBER: ${{ github.event.number }}
+          PORTER_PROJECT: "12565"
+          PORTER_REPO_NAME: ${{ github.event.repository.name }}
+          PORTER_STACK_NAME: worker
+          PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
+          PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4060 }}

.github/workflows/porter_stack_worker_us.yml

@@ -1,29 +0,0 @@
-"on":
-  push:
-    branches:
-      - production
-name: Deploy to worker (US)
-jobs:
-  porter-deploy:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout code
-        uses: actions/checkout@v3
-      - name: Set Github tag
-        id: vars
-        run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
-      - name: Setup porter
-        uses: porter-dev/setup-porter@v0.1.0
-      - name: Deploy stack
-        timeout-minutes: 30
-        run: exec porter apply
-        env:
-          PORTER_CLUSTER: "4038"
-          PORTER_DEPLOYMENT_TARGET_ID: 260cd04e-5dc8-4b22-a291-ad4453f981f7
-          PORTER_HOST: https://dashboard.getporter.dev
-          PORTER_PR_NUMBER: ${{ github.event.number }}
-          PORTER_PROJECT: "12565"
-          PORTER_REPO_NAME: ${{ github.event.repository.name }}
-          PORTER_STACK_NAME: worker
-          PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
-          PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4038 }}

.github/workflows/snyk.yml

@@ -1,5 +1,12 @@
 name: Snyk Container
-on: push
+on:
+  pull_request:
+    branches:
+      - "**"
+  push:
+    branches:
+      - main
+
 jobs:
   snyk:
     runs-on: ubuntu-latest
@@ -18,7 +25,7 @@ jobs:
           args: --file=web/Dockerfile
 
       - name: Upload result to GitHub Code Scanning
-        uses: github/codeql-action/upload-sarif@v2
+        uses: github/codeql-action/upload-sarif@v3
         with:
           sarif_file: snyk.sarif
           category: web
@@ -33,7 +40,7 @@ jobs:
           args: --file=worker/Dockerfile
 
       - name: Upload result to GitHub Code Scanning
-        uses: github/codeql-action/upload-sarif@v2
+        uses: github/codeql-action/upload-sarif@v3
         with:
           sarif_file: snyk.sarif
           category: worker

CONTRIBUTING.md

@@ -53,9 +53,7 @@ A good first step is to search for open [issues](https://github.com/langfuse/lan
 
 ### Architecture Overview
 
-> [!NOTE]
-> Infrastructure will change in Langfuse version 3.0. More in the [GitHub Discussions](https://github.com/orgs/langfuse/discussions/1902).
-> `langfuse/langfuse/worker` is under active development and not recommended for production use in Langfuse 2.x.
+**Langfuse v2**
 
 ```mermaid
 flowchart TB
@@ -93,15 +91,66 @@ flowchart TB
         I --- Prisma
     end
 
-    subgraph s5["Application (langfuse/langfuse/worker)"]
-        Worker_API[Public HTTP API]
+    Prisma --- DB
+    JS --- API
+    Python --- API
+```
+
+**Langfuse v3 (work in progress, not released yet)**
+
+> [!NOTE]
+> Infrastructure will change in Langfuse version 3.0. More in the [GitHub Discussions](https://github.com/orgs/langfuse/discussions/1902).
+> `langfuse/langfuse/worker` is under active development and not recommended for production use in Langfuse 2.x.
+
+```mermaid
+flowchart TB
+    subgraph s4["Clients"]
+        subgraph s2["langfuse/langfuse-python"]
+            Python["Python low-level SDK"]
+            Decorator["observe() decorator"] -->|extends| Python
+            OAI["OpenAI drop-in replacement"] -->|extends| Python
+            Llamaindex["LlamaIndex Integration"] -->|extends| Python
+            LCPYTHON["Langchain Python Integration"] -->|extends| Python
+            Langflow -->|uses| LCPYTHON
+            LiteLLM -->|uses| Python
+        end
+        subgraph s3["langfuse/langfuse-js"]
+            JS["JS SDK"]
+            LCJS["Langchain JS Integration"]  -->|extends| JS
+            Flowise -->|uses| LCJS
+        end
     end
 
-    API --> Worker_API
-    Worker_API --- DB
-    Worker_API --- Redis
+    DB[Postgres Database]
+    Redis[Redis Cache/Queue]
+    Clickhouse[Clickhouse Database]
+
+    subgraph s1["Application (langfuse/langfuse/web)"]
+        API[Public HTTP API]
+        G[TRPC API]
+        I[NextAuth]
+        H[React Frontend]
+        ORM
+        H --> G
+        H --> I
+        G --> I
+        G --- ORM
+        API --- ORM
+        I --- ORM
+    end
+
+    subgraph s5["Application (langfuse/langfuse/worker)"]
+        Worker
+    end
+
+    Worker --- DB
+    Worker --- Redis
+    Worker --- Clickhouse
+
+    ORM --- DB
+    ORM --- Redis
+    ORM --- Clickhouse
 
-    Prisma --- DB
     JS --- API
     Python --- API
 ```
@@ -116,7 +165,7 @@ flowchart LR
    Browser ---|Web UI & TRPC API| App
    Integrations/SDKs ---|Public HTTP API| App
    subgraph i1["Application Network"]
-      App["Langfuse Application (Docker or Serverless)"]
+      App["Langfuse Application"]
    end
    subgraph i2["Database Network"]
       DB["Postgres Database"]
@@ -149,6 +198,7 @@ We built a monorepo using [pnpm](https://pnpm.io/motivation) and [turbo](https:/
 Requirements
 
 - Node.js 20 as specified in the [.nvmrc](.nvmrc)
+- Pnpm v.9.5.0
 - Docker to run the database locally
 
 **Note:** You can also simply run Langfuse in a **GitHub Codespace** via the provided devcontainer. To do this, click on the green "Code" button in the top right corner of the repository and select "Open with Codespaces".

README.md

@@ -1,4 +1,8 @@
-![Langfuse GitHub Banner](https://github.com/langfuse/langfuse/assets/121163007/6035f0f3-d691-4963-b5d0-10cf506e9d42)
+![langfuse_logo_1024](https://github.com/user-attachments/assets/f17d8fc1-8206-4afa-a5ec-774826be88a8)
+
+<div align="center"><h1> Langfuse: Open Source LLM Observability & Engineering Platform</h4></div>
+<div align="center"><h2> Debug and improve your LLM app</h4></div>
+<div align="center"><h3> LLM Observability, Prompt Management, LLM Evaluations, Datasets, LLM Metrics and Prompt Playground</h4></div>
 
 <div align="center">
    <div>
@@ -37,7 +41,7 @@
 </div>
 </br>
 
-## Overview
+## Langfuse Overview
 
 _Unmute video for voice-over_
 
@@ -45,18 +49,18 @@ https://github.com/langfuse/langfuse/assets/2834609/a94062e9-c782-4ee9-af59-dee6
 
 ### Develop
 
-- **Observability:** Instrument your app and start ingesting traces to Langfuse ([Quickstart](https://langfuse.com/docs/get-started), [Integrations](https://langfuse.com/docs/integrations) [Tracing](https://langfuse.com/docs/tracing))
+- **LLM Observability:** Instrument your app and start ingesting traces to Langfuse ([Quickstart](https://langfuse.com/docs/get-started), [Integrations](https://langfuse.com/docs/integrations) [Tracing](https://langfuse.com/docs/tracing))
 - **Langfuse UI:** Inspect and debug complex logs ([Demo](https://langfuse.com/docs/demo), [Tracing](https://langfuse.com/docs/tracing))
 - **Prompt Management:** Manage, version and deploy prompts from within Langfuse ([Prompt Management](https://langfuse.com/docs/prompts/get-started))
 - **Prompt Engineering:** Test and iterate on your prompts with the [LLM Playground](https://langfuse.com/docs/playground)
 
 ### Monitor
 
-- **Analytics:** Track metrics (cost, latency, quality) and gain insights from dashboards & data exports ([Analytics](https://langfuse.com/docs/analytics))
-- **Evals:** Collect and calculate scores for your LLM completions ([Scores & Evaluations](https://langfuse.com/docs/scores))
-  - Run model-based evaluations ([Model-based evaluations](https://langfuse.com/docs/scores/model-based-evals)) within Langfuse
+- **LLM Analytics:** Track metrics (cost, latency, quality) and gain insights from dashboards & data exports ([Analytics](https://langfuse.com/docs/analytics))
+- **LLM Evaluations:** Collect and calculate scores for your LLM completions ([Scores & Evaluations](https://langfuse.com/docs/scores))
+  - Run ([Model-based evaluations](https://langfuse.com/docs/scores/model-based-evals)) and LLM-as-a-Judge within Langfuse
   - Collect user feedback ([User Feedback](https://langfuse.com/docs/scores/user-feedback))
-  - Manually score observations in Langfuse ([Manual Scores](https://langfuse.com/docs/scores/manually))
+  - Manually score LLM outputs in Langfuse ([Manual Scores](https://langfuse.com/docs/scores/manually))
 
 ### Test
 
@@ -72,6 +76,8 @@ Managed deployment by the Langfuse team, generous free-tier (hobby plan), no cre
 
 **[» Langfuse Cloud](https://cloud.langfuse.com)**
 
+## Self-Hosting Open Source LLM Observability with Langfuse
+
 ### Localhost (docker)
 
 ```bash
@@ -98,7 +104,7 @@ Templated deployments: [Railway, GCP Cloud Run, AWS Fargate, Kubernetes and othe
 
 You need a Langfuse public and secret key to get started. Sign up [here](https://cloud.langfuse.com) and find them in your project settings.
 
-### Ingesting Data · Instrumenting Your Application
+### Ingesting Data · Instrumenting Your Application · LLM Observability with Langfuse
 
 Note: We recommend using our fully async, typed [SDKs](https://langfuse.com/docs/sdk) that allow you to instrument any LLM application with any underlying model. They are available in [Python (Decorators)](https://langfuse.com/docs/sdk/python) & [JS/TS](https://langfuse.com/docs/sdk/typescript). The SDKs will always be the most fully featured and stable way to ingest data into Langfuse.
 
@@ -106,7 +112,7 @@ You may want to use another integration to get started quickly or implement a us
 
 See the [→ Quickstart](https://langfuse.com/docs/get-started) to integrate Langfuse.
 
-### Integrations
+### LLM Observability Integrations
 
 | Integration                                              | Supports                   | Description                                                                                                                                      |
 | -------------------------------------------------------- | -------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------ |
@@ -127,15 +133,14 @@ Packages that integrate with Langfuse:
 | [AI SDK by Vercel](/docs/sdk/typescript/example-vercel-ai) | Typescript SDK that makes streaming LLM outputs super easy.                                                             |
 | [Flowise](/docs/integrations/flowise)                      | JS/TS no-code builder for customized LLM flows.                                                                         |
 | [Langflow](/docs/integrations/langflow)                    | Python-based UI for LangChain, designed with react-flow to provide an effortless way to experiment and prototype flows. |
-| [Superagent](/docs/integrations/superagent)                | Open Source AI Assistant Framework & API for prototyping and deployment of agents.                                      |
 
 ## Questions and feedback
 
 ### Ideas and roadmap
 
-- [GitHub Discussions](https://github.com/orgs/langfuse/discussions)
-- [Feature Requests](https://langfuse.com/idea)
 - [Roadmap](https://langfuse.com/roadmap)
+- [GitHub Discussions](https://github.com/orgs/langfuse/discussions)
+- [Feature Requests](https://langfuse.com/idea
 
 ### Support and feedback
 

careers

@@ -3,7 +3,6 @@ Join us in scaling Langfuse in Berlin, Germany. We are an open source company, w
 
 _Open Roles_
 
-- Backend Engineer, 70-130k EUR, 0.25-0.75% Equity, https://www.ycombinator.com/companies/langfuse/jobs/mnrdwla-backend-engineer
 - Product Engineer, 70-130k EUR, 0.25-0.75% Equity, https://www.ycombinator.com/companies/langfuse/jobs/aAvmoFB-product-engineer
 - Developer Advocate, 60-110k EUR, 0.25-0.5% Equity, https://www.ycombinator.com/companies/langfuse/jobs/uHysbKH-developer-advocate-devrel
 

docker-compose.build.yml

@@ -16,6 +16,7 @@ services:
       - DATABASE_URL=postgresql://postgres:postgres@db:5432/postgres
       - NEXTAUTH_SECRET=mysecret
       - SALT=mysalt
+      - ENCRYPTION_KEY=0000000000000000000000000000000000000000000000000000000000000000 # generate via `openssl rand -hex 32`
       - NEXTAUTH_URL=http://localhost:3000
       - TELEMETRY_ENABLED=${TELEMETRY_ENABLED:-true}
       - LANGFUSE_ENABLE_EXPERIMENTAL_FEATURES=${LANGFUSE_ENABLE_EXPERIMENTAL_FEATURES:-false}

docker-compose.dev.yml

@@ -1,4 +1,22 @@
 services:
+  clickhouse:
+    image: clickhouse/clickhouse-server
+    user: "101:101"
+    container_name: clickhouse
+    hostname: clickhouse
+    environment:
+      CLICKHOUSE_DB: default
+      CLICKHOUSE_USER: clickhouse
+      CLICKHOUSE_PASSWORD: clickhouse
+    volumes:
+      - langfuse_clickhouse_data:/var/lib/clickhouse
+      - langfuse_clickhouse_logs:/var/log/clickhouse-server
+    ports:
+      - "8123:8123"
+      - "9000:9000"
+    depends_on:
+      - postgres
+
   redis:
     image: redis:7.2.4
     restart: always
@@ -7,7 +25,7 @@ services:
     ports:
       - 6379:6379
 
-  db:
+  postgres:
     image: postgres:${POSTGRES_VERSION:-latest}
     restart: always
     command: ["postgres", "-c", "log_statement=all"]
@@ -18,8 +36,12 @@ services:
     ports:
       - 5432:5432
     volumes:
-      - database_data:/var/lib/postgresql/data
+      - langfuse_postgres_data:/var/lib/postgresql/data
 
 volumes:
-  database_data:
+  langfuse_postgres_data:
+    driver: local
+  langfuse_clickhouse_data:
+    driver: local
+  langfuse_clickhouse_logs:
     driver: local

docker-compose.yml

@@ -10,6 +10,7 @@ services:
       - DATABASE_URL=postgresql://postgres:postgres@db:5432/postgres
       - NEXTAUTH_SECRET=mysecret
       - SALT=mysalt
+      - ENCRYPTION_KEY=0000000000000000000000000000000000000000000000000000000000000000 # generate via `openssl rand -hex 32`
       - NEXTAUTH_URL=http://localhost:3000
       - TELEMETRY_ENABLED=${TELEMETRY_ENABLED:-true}
       - LANGFUSE_ENABLE_EXPERIMENTAL_FEATURES=${LANGFUSE_ENABLE_EXPERIMENTAL_FEATURES:-false}

ee/package.json

@@ -15,23 +15,24 @@
     }
   },
   "engines": {
-    "node": ">=20"
+    "node": "20"
   },
   "scripts": {
     "build": "tsc",
     "dev": "tsc --watch",
-    "lint": "eslint . --ext .js,.jsx,.ts,.tsx",
+    "lint": "eslint . --ext .js,.jsx,.ts,.tsx --max-warnings 0",
     "lint:fix": "eslint . --ext .js,.jsx,.ts,.tsx --fix"
   },
   "dependencies": {
     "@langfuse/shared": "workspace:*",
-    "next": "^14.2.3",
+    "@opentelemetry/api": ">=1.0.0 <1.9.0",
+    "next": "^14.2.6",
     "next-auth": "^4.24.7",
     "zod": "^3.23.8"
   },
   "devDependencies": {
-    "@repo/eslint-config": "*",
-    "@repo/typescript-config": "*",
+    "@repo/eslint-config": "workspace:*",
+    "@repo/typescript-config": "workspace:*",
     "@types/node": "^20.11.29",
     "@typescript-eslint/parser": "^7.12.0",
     "eslint": "^8.57.0",

ee/src/ee-license-check/index.ts

@@ -1,4 +1,5 @@
 import { env } from "../env";
 
 export const isEeAvailable: boolean =
-  env.NEXT_PUBLIC_LANGFUSE_CLOUD_REGION !== undefined;
+  env.NEXT_PUBLIC_LANGFUSE_CLOUD_REGION !== undefined ||
+  env.LANGFUSE_EE_LICENSE_KEY !== undefined;

ee/src/env.ts

@@ -3,7 +3,7 @@ import { env as sharedEnv } from "@langfuse/shared";
 
 const EnvSchema = z.object({
   NEXT_PUBLIC_LANGFUSE_CLOUD_REGION: z.string().optional(),
-  ADMIN_API_KEY: z.string().optional(),
+  LANGFUSE_EE_LICENSE_KEY: z.string().optional(),
 });
 
 export const env = { ...sharedEnv, ...EnvSchema.parse(process.env) };

ee/src/index.ts

@@ -1 +0,0 @@
-export { isEeAvailable } from "./ee-license-check";

fern/apis/server/definition/commons.yml

@@ -5,16 +5,36 @@ types:
       id:
         type: string
         docs: The unique identifier of a trace
-      timestamp: datetime
-      name: optional<string>
-      input: optional<unknown>
-      output: optional<unknown>
-      sessionId: optional<string>
-      release: optional<string>
-      version: optional<string>
-      userId: optional<string>
-      metadata: optional<unknown>
-      tags: optional<list<string>>
+      timestamp:
+        type: datetime
+        docs: The timestamp when the trace was created
+      name:
+        type: optional<string>
+        docs: The name of the trace
+      input:
+        type: optional<unknown>
+        docs: The input data of the trace. Can be any JSON.
+      output:
+        type: optional<unknown>
+        docs: The output data of the trace. Can be any JSON.
+      sessionId:
+        type: optional<string>
+        docs: The session identifier associated with the trace
+      release:
+        type: optional<string>
+        docs: The release version of the application when the trace was created
+      version:
+        type: optional<string>
+        docs: The version of the trace
+      userId:
+        type: optional<string>
+        docs: The user identifier associated with the trace
+      metadata:
+        type: optional<unknown>
+        docs: The metadata associated with the trace. Can be any JSON.
+      tags:
+        type: optional<list<string>>
+        docs: The tags associated with the trace. Can be an array of strings or null.
       public:
         type: optional<boolean>
         docs: Public traces are accessible via url without login
@@ -65,37 +85,97 @@ types:
       traces: list<Trace>
   Observation:
     properties:
-      id: string
-      traceId: optional<string>
-      type: string
-      name: optional<string>
-      startTime: datetime
-      endTime: optional<datetime>
-      completionStartTime: optional<datetime>
-      model: optional<string>
-      modelParameters: optional<map<string, MapValue>>
-      input: optional<unknown>
-      version: optional<string>
-      metadata: optional<unknown>
-      output: optional<unknown>
-      usage: optional<Usage>
-      level: ObservationLevel
-      statusMessage: optional<string>
-      parentObservationId: optional<string>
-      promptId: optional<string>
+      id:
+        type: string
+        docs: The unique identifier of the observation
+      traceId:
+        type: optional<string>
+        docs: The trace ID associated with the observation
+      type:
+        type: string
+        docs: The type of the observation
+      name:
+        type: optional<string>
+        docs: The name of the observation
+      startTime:
+        type: datetime
+        docs: The start time of the observation
+      endTime:
+        type: optional<datetime>
+        docs: The end time of the observation.
+      completionStartTime:
+        type: optional<datetime>
+        docs: The completion start time of the observation
+      model:
+        type: optional<string>
+        docs: The model used for the observation
+      modelParameters:
+        type: optional<map<string, MapValue>>
+        docs: The parameters of the model used for the observation
+      input:
+        type: optional<unknown>
+        docs: The input data of the observation
+      version:
+        type: optional<string>
+        docs: The version of the observation
+      metadata:
+        type: optional<unknown>
+        docs: Additional metadata of the observation
+      output:
+        type: optional<unknown>
+        docs: The output data of the observation
+      usage:
+        type: optional<Usage>
+        docs: The usage data of the observation
+      level:
+        type: ObservationLevel
+        docs: The level of the observation
+      statusMessage:
+        type: optional<string>
+        docs: The status message of the observation
+      parentObservationId:
+        type: optional<string>
+        docs: The parent observation ID
+      promptId:
+        type: optional<string>
+        docs: The prompt ID associated with the observation
 
   ObservationsView:
     extends: Observation
     properties:
-      modelId: optional<string>
-      inputPrice: optional<double>
-      outputPrice: optional<double>
-      totalPrice: optional<double>
-      calculatedInputCost: optional<double>
-      calculatedOutputCost: optional<double>
-      calculatedTotalCost: optional<double>
-      latency: optional<double>
-      timeToFirstToken: optional<double>
+      promptName:
+        type: optional<string>
+        docs: The name of the prompt associated with the observation
+      promptVersion:
+        type: optional<integer>
+        docs: The version of the prompt associated with the observation
+      modelId:
+        type: optional<string>
+        docs: The unique identifier of the model
+      inputPrice:
+        type: optional<double>
+        docs: The price of the input in USD
+      outputPrice:
+        type: optional<double>
+        docs: The price of the output in USD.
+      totalPrice:
+        type: optional<double>
+        docs: The total price in USD.
+      calculatedInputCost:
+        type: optional<double>
+        docs: The calculated cost of the input in USD
+      calculatedOutputCost:
+        type: optional<double>
+        docs: The calculated cost of the output in USD
+      calculatedTotalCost:
+        type: optional<double>
+        docs: The calculated total cost in USD
+      latency:
+        type: optional<double>
+        docs: The latency in seconds.
+      timeToFirstToken:
+        type: optional<double>
+        docs: The time to the first token in seconds
 
   Usage:
     docs: Standard interface for usage and cost
@@ -237,14 +317,30 @@ types:
       updatedAt: datetime
   DatasetRun:
     properties:
-      id: string
-      name: string
-      description: optional<string>
-      metadata: optional<unknown>
-      datasetId: string
-      datasetName: string
-      createdAt: datetime
-      updatedAt: datetime
+      id:
+        type: string
+        docs: Unique identifier of the dataset run
+      name:
+        type: string
+        docs: Name of the dataset run
+      description:
+        type: optional<string>
+        docs: Description of the run
+      metadata:
+        type: optional<unknown>
+        docs: Metadata of the dataset run
+      datasetId:
+        type: string
+        docs: Id of the associated dataset
+      datasetName:
+        type: string
+        docs: Name of the associated dataset
+      createdAt:
+        type: datetime
+        docs: The date and time when the dataset run was created
+      updatedAt:
+        type: datetime
+        docs: The date and time when the dataset run was last updated
   DatasetRunWithItems:
     extends: DatasetRun
     properties:
@@ -292,6 +388,7 @@ types:
       - MILLISECONDS
       - SECONDS
       - IMAGES
+      - REQUESTS
   ObservationLevel:
     enum:
       - DEBUG

fern/apis/server/definition/dataset-items.yml

@@ -49,7 +49,7 @@ types:
       sourceObservationId: optional<string>
       id:
         type: optional<string>
-        docs: Dataset items are upserted on their id. Id needs to be globally unique and cannot be reused across datasets.
+        docs: Dataset items are upserted on their id. Id needs to be unique (project-level) and cannot be reused across datasets.
       status:
         type: optional<commons.DatasetStatus>
         docs: Defaults to ACTIVE for newly created items

fern/apis/server/definition/observations.yml

@@ -39,6 +39,9 @@ service:
           toStartTime:
             type: optional<datetime>
             docs: Retrieve only observations with a start_time before this datetime (ISO 8601).
+          version:
+            type: optional<string>
+            docs: Optional filter to only include observations with a certain version.
       response: ObservationsViews
 
 types:

fern/apis/server/definition/prompts.yml

@@ -41,11 +41,17 @@ service:
           limit:
             type: optional<integer>
             docs: limit of items per page
+          fromUpdatedAt:
+            type: optional<datetime>
+            docs: Optional filter to only include prompt versions created/updated on or after a certain datetime (ISO 8601)
+          toUpdatedAt:
+            type: optional<datetime>
+            docs: Optional filter to only include prompt versions created/updated before a certain datetime (ISO 8601)
 
       response: PromptMetaListResponse
 
     create:
-      docs: Create a prompt
+      docs: Create a new version for the prompt with the given `name`
       method: POST
       path: /prompts
       request: CreatePromptRequest
@@ -63,6 +69,10 @@ types:
       versions: list<integer>
       labels: list<string>
       tags: list<string>
+      lastUpdatedAt: datetime
+      lastConfig:
+        type: unknown
+        docs: Config object of the most recent prompt version that matches the filters (if any are provided)
 
   CreatePromptRequest:
     union:

fern/apis/server/definition/sessions.yml

@@ -7,7 +7,7 @@ service:
   base-path: /api/public
   endpoints:
     list:
-      docs: Get sessions.
+      docs: Get sessions
       method: GET
       path: /sessions
       request:

fern/apis/server/definition/trace.yml

@@ -16,7 +16,7 @@ service:
           docs: The unique langfuse identifier of a trace
       response: commons.TraceWithFullDetails
     list:
-      docs: Get list of traces.
+      docs: Get list of traces
       method: GET
       path: /traces
       request:
@@ -44,6 +44,12 @@ service:
             type: optional<string>
             allow-multiple: true
             docs: Only traces that include all of these tags will be returned.
+          version:
+            type: optional<string>
+            docs: Optional filter to only include traces with a certain version.
+          release:
+            type: optional<string>
+            docs: Optional filter to only include traces with a certain release.
       response: Traces
 
 types:

package.json

@@ -1,12 +1,11 @@
 {
   "name": "langfuse",
-  "version": "2.63.1",
+  "version": "2.78.0",
   "author": "engineering@langfuse.com",
   "license": "MIT",
   "private": true,
   "engines": {
-    "node": ">=20",
-    "pnpm": "<9"
+    "node": "20"
   },
   "scripts": {
     "preinstall": "npx only-allow pnpm",
@@ -17,8 +16,9 @@
     "db:seed": "turbo run db:seed",
     "db:seed:examples": "turbo run db:seed:examples",
     "nuke": "bash ./scripts/nuke.sh",
-    "dx": "pnpm i && pnpm run infra:dev:up && pnpm --filter=shared run db:reset && pnpm --filter=shared run db:seed:examples && pnpm run dev",
-    "dx-f": "pnpm i && pnpm run infra:dev:up && pnpm --filter=shared run db:reset -f && pnpm --filter=shared run db:seed:examples && pnpm run dev",
+    "dx": "pnpm i && pnpm run infra:dev:up && pnpm --filter=shared run db:reset && pnpm --filter=shared run ch:reset && pnpm --filter=shared run db:seed:examples && pnpm run dev",
+    "dx-f": "pnpm i && pnpm run infra:dev:up && pnpm --filter=shared run db:reset -f && pnpm --filter=shared run ch:reset && pnpm --filter=shared run db:seed:examples && pnpm run dev",
+    "dx:skip-infra": "pnpm i && pnpm --filter=shared run db:reset && pnpm --filter=shared run ch:reset && pnpm --filter=shared run db:seed:examples && pnpm run dev",
     "build": "turbo run build",
     "start": "turbo run start",
     "dev": "turbo run dev",
@@ -35,7 +35,7 @@
     "husky": "^9.0.11",
     "prettier": "^3.2.5",
     "release-it": "^17.3.0",
-    "turbo": "^1.13.3"
+    "turbo": "^1.13.4"
   },
   "release-it": {
     "git": {
@@ -79,5 +79,6 @@
         "pr": ":rocket: _This pull request is included in v${version}. See [${releaseName}](${releaseUrl}) for release notes._"
       }
     }
-  }
+  },
+  "packageManager": "pnpm@9.5.0"
 }

packages/config-eslint/package.json

@@ -11,9 +11,9 @@
     "@typescript-eslint/eslint-plugin": "^7.1.0",
     "@typescript-eslint/parser": "^7.12.0",
     "@vercel/style-guide": "^6.0.0",
-    "eslint-config-next": "^14.2.3",
+    "eslint-config-next": "^14.2.6",
     "eslint-config-prettier": "^9.1.0",
-    "eslint-config-turbo": "^1.13.3",
+    "eslint-config-turbo": "^1.13.4",
     "eslint-plugin-only-warn": "^1.1.0",
     "typescript": "^5.4.5"
   }

packages/shared/clickhouse/migrations/0001_traces.down.sql

@@ -0,0 +1 @@
+DROP TABLE traces;

packages/shared/clickhouse/migrations/0001_traces.up.sql

@@ -0,0 +1,26 @@
+CREATE TABLE traces (
+    `id` String,
+    `timestamp` DateTime64(3),
+    `name` String,
+    `user_id` Nullable(String),
+    `metadata` Map(String, String) CODEC(ZSTD(1)),
+    `release` Nullable(String),
+    `version` Nullable(String),
+    `project_id` String,
+    `public` Bool,
+    `bookmarked` Bool,
+    `tags` Array(String),
+    `input` Nullable(String) CODEC(ZSTD(1)),
+    `output` Nullable(String) CODEC(ZSTD(1)),
+    `session_id` Nullable(String),
+    `created_at` DateTime64(3) DEFAULT now(),
+    `updated_at` DateTime64(3) DEFAULT now(),
+    INDEX idx_id id TYPE bloom_filter(0.001) GRANULARITY 1,
+    INDEX idx_res_metadata_key mapKeys(metadata) TYPE bloom_filter(0.01) GRANULARITY 1,
+    INDEX idx_res_metadata_value mapValues(metadata) TYPE bloom_filter(0.01) GRANULARITY 1
+) ENGINE = ReplacingMergeTree Partition by toYYYYMM(timestamp)
+ORDER BY (
+        project_id,
+        toUnixTimestamp(timestamp),
+        id
+    );

packages/shared/clickhouse/migrations/0002_observations.down.sql

@@ -0,0 +1 @@
+DROP TABLE observations;

packages/shared/clickhouse/migrations/0002_observations.up.sql

@@ -0,0 +1,50 @@
+CREATE TABLE observations (
+    `id` String,
+    `trace_id` String,
+    `project_id` String,
+    `type` LowCardinality(String),
+    `parent_observation_id` Nullable(String),
+    `start_time` DateTime64(3),
+    `end_time` Nullable(DateTime64(3)),
+    `name` String,
+    `metadata` Map(LowCardinality(String), String) CODEC(ZSTD(1)),
+    `level` LowCardinality(String),
+    `status_message` Nullable(String),
+    `version` Nullable(String),
+    `input` Nullable(String) CODEC(ZSTD(3)),
+    `output` Nullable(String) CODEC(ZSTD(3)),
+    `provided_model_name` Nullable(String),
+    `internal_model_id` Nullable(String),
+    `model_parameters` Nullable(String),
+    `provided_input_usage_units` Nullable(Decimal64(12)),
+    `provided_output_usage_units` Nullable(Decimal64(12)),
+    `provided_total_usage_units` Nullable(Decimal64(12)),
+    `input_usage_units` Nullable(Decimal64(12)),
+    `output_usage_units` Nullable(Decimal64(12)),
+    `total_usage_units` Nullable(Decimal64(12)),
+    `unit` Nullable(String),
+    `provided_input_cost` Nullable(Decimal64(12)),
+    `provided_output_cost` Nullable(Decimal64(12)),
+    `provided_total_cost` Nullable(Decimal64(12)),
+    `input_cost` Nullable(Decimal64(12)),
+    `output_cost` Nullable(Decimal64(12)),
+    `total_cost` Nullable(Decimal64(12)),
+    `completion_start_time` Nullable(DateTime64(3)),
+    `prompt_id` Nullable(String),
+    `prompt_name` Nullable(String),
+    `prompt_version` Nullable(UInt16),
+    `created_at` DateTime64(3) DEFAULT now(),
+    `updated_at` DateTime64(3) DEFAULT now(),
+    INDEX idx_id id TYPE bloom_filter() GRANULARITY 1,
+    INDEX idx_trace_id trace_id TYPE bloom_filter() GRANULARITY 1,
+    INDEX idx_project_id project_id TYPE bloom_filter() GRANULARITY 1,
+    INDEX idx_res_metadata_key mapKeys(metadata) TYPE bloom_filter() GRANULARITY 1,
+    INDEX idx_res_metadata_value mapValues(metadata) TYPE bloom_filter() GRANULARITY 1
+) ENGINE = ReplacingMergeTree Partition by toYYYYMM(start_time)
+ORDER BY (
+        project_id,
+        `type`,
+        trace_id,
+        toUnixTimestamp(start_time),
+        id
+    );

packages/shared/clickhouse/migrations/0003_scores.down.sql

@@ -0,0 +1 @@
+DROP TABLE scores;

packages/shared/clickhouse/migrations/0003_scores.up.sql

@@ -0,0 +1,25 @@
+CREATE TABLE scores (
+    `id` String,
+    `timestamp` DateTime64(3),
+    `project_id` String,
+    `trace_id` String,
+    `observation_id` Nullable(String),
+    `name` String,
+    `value` Float64,
+    `source` String,
+    `comment` Nullable(String) CODEC(ZSTD(1)),
+    `author_user_id` Nullable(String),
+    `config_id` Nullable(String),
+    `data_type` String,
+    `string_value` Nullable(String),
+    `created_at` DateTime64(3) DEFAULT now(),
+    `updated_at` DateTime64(3) DEFAULT now(),
+    INDEX idx_id id TYPE bloom_filter(0.001) GRANULARITY 1,
+    INDEX idx_project_id trace_id TYPE bloom_filter(0.001) GRANULARITY 1
+) ENGINE = ReplacingMergeTree Partition by toYYYYMM(timestamp)
+ORDER BY (
+        project_id,
+        trace_id,
+        toUnixTimestamp(timestamp),
+        id
+    );

packages/shared/clickhouse/scripts/down.sh

@@ -0,0 +1,22 @@
+#!/bin/bash
+
+# Load environment variables
+source ../../.env
+
+# Check if golang-migrate is installed
+if ! command -v migrate &> /dev/null
+then
+    echo "Error: golang-migrate is not installed or not in PATH."
+    echo "Please install golang-migrate via 'brew install golang-migrate' to run this script."
+    echo "Visit https://github.com/golang-migrate/migrate for more installation instructions."
+    exit 1
+fi
+
+# Construct the database URL
+if [ "$CLICKHOUSE_MIGRATION_SSL" = true ] ; then
+    DATABASE_URL="${CLICKHOUSE_MIGRATION_URL}?username=${CLICKHOUSE_USER}&password=${CLICKHOUSE_PASSWORD}&database=default&x-multi-statement=true&secure=true&skip_verify=true&x-migrations-table-engine=MergeTree"
+else
+    DATABASE_URL="${CLICKHOUSE_MIGRATION_URL}?username=${CLICKHOUSE_USER}&password=${CLICKHOUSE_PASSWORD}&database=default&x-multi-statement=true&x-migrations-table-engine=MergeTree"
+fi
+# Execute the down command
+migrate -source file://clickhouse/migrations -database "$DATABASE_URL" down

packages/shared/clickhouse/scripts/drop.sh

@@ -0,0 +1,22 @@
+#!/bin/bash
+
+# Load environment variables
+source ../../.env
+
+# Check if golang-migrate is installed
+if ! command -v migrate &> /dev/null
+then
+    echo "Error: golang-migrate is not installed or not in PATH."
+    echo "Please install golang-migrate via 'brew install golang-migrate' to run this script."
+    echo "Visit https://github.com/golang-migrate/migrate for more installation instructions."
+    exit 1
+fi
+
+# Construct the database URL
+if [ "$CLICKHOUSE_MIGRATION_SSL" = true ] ; then
+    DATABASE_URL="${CLICKHOUSE_MIGRATION_URL}?username=${CLICKHOUSE_USER}&password=${CLICKHOUSE_PASSWORD}&database=default&x-multi-statement=true&secure=true&skip_verify=true&x-migrations-table-engine=MergeTree"
+else
+    DATABASE_URL="${CLICKHOUSE_MIGRATION_URL}?username=${CLICKHOUSE_USER}&password=${CLICKHOUSE_PASSWORD}&database=default&x-multi-statement=true&x-migrations-table-engine=MergeTree"
+fi
+# Execute the drop command
+migrate -source file://clickhouse/migrations -database "$DATABASE_URL" drop

packages/shared/clickhouse/scripts/up.sh

@@ -0,0 +1,23 @@
+#!/bin/bash
+
+# Load environment variables
+source ../../.env
+
+# Check if golang-migrate is installed
+if ! command -v migrate &> /dev/null
+then
+    echo "Error: golang-migrate is not installed or not in PATH."
+    echo "Please install golang-migrate via 'brew install golang-migrate' to run this script."
+    echo "Visit https://github.com/golang-migrate/migrate for more installation instructions."
+    exit 1
+fi
+
+# Construct the database URL
+if [ "$CLICKHOUSE_MIGRATION_SSL" = true ] ; then
+    DATABASE_URL="${CLICKHOUSE_MIGRATION_URL}?username=${CLICKHOUSE_USER}&password=${CLICKHOUSE_PASSWORD}&database=default&x-multi-statement=true&secure=true&skip_verify=true&x-migrations-table-engine=MergeTree"
+else
+    DATABASE_URL="${CLICKHOUSE_MIGRATION_URL}?username=${CLICKHOUSE_USER}&password=${CLICKHOUSE_PASSWORD}&database=default&x-multi-statement=true&x-migrations-table-engine=MergeTree"
+fi
+
+# Execute the up command
+migrate -source file://clickhouse/migrations -database "$DATABASE_URL" up

packages/shared/package.json

@@ -6,7 +6,7 @@
   "main": "./dist/src/index.js",
   "types": "./dist/src/index.d.ts",
   "engines": {
-    "node": ">=20"
+    "node": "20"
   },
   "exports": {
     ".": {
@@ -21,6 +21,10 @@
       "import": "./dist/src/server/index.js",
       "require": "./dist/src/server/index.js"
     },
+    "./src/server/auth/apiKeys": {
+      "import": "./dist/src/server/auth/apiKeys.js",
+      "require": "./dist/src/server/auth/apiKeys.js"
+    },
     "./encryption": {
       "import": "./dist/src/encryption/index.js",
       "require": "./dist/src/encryption/index.js"
@@ -38,36 +42,51 @@
     "db:seed": "dotenv -e ../../.env -- npx prisma db seed",
     "db:generate": "dotenv -e ../../.env -- npx prisma generate",
     "db:seed:examples": "dotenv -e ../../.env -- npx prisma db seed -- --environment examples",
-    "db:seed:load": "dotenv -e ../../.env -- npx prisma db seed -- --environment load"
+    "db:seed:load": "dotenv -e ../../.env -- npx prisma db seed -- --environment load",
+    "ch:status": "dotenv -e ../../.env -- goose -dir './clickhouse/migrations/' status",
+    "ch:up": "bash clickhouse/scripts/up.sh",
+    "ch:down": "bash clickhouse/scripts/down.sh",
+    "ch:drop": "bash clickhouse/scripts/drop.sh",
+    "ch:reset": "pnpm run ch:down && pnpm run ch:up"
   },
   "prisma": {
     "seed": "ts-node -r tsconfig-paths/register -r dotenv/config --compiler-options {\"module\":\"CommonJS\"} prisma/seed.ts"
   },
   "dependencies": {
-    "@aws-sdk/client-s3": "^3.550.0",
+    "@anthropic-ai/tokenizer": "^0.0.4",
+    "@aws-sdk/client-s3": "^3.627.0",
     "@aws-sdk/lib-storage": "^3.568.0",
     "@aws-sdk/s3-request-presigner": "^3.554.0",
+    "@clickhouse/client": "^1.4.0",
     "@langchain/anthropic": "^0.1.21",
     "@langchain/core": "^0.1.61",
     "@langchain/openai": "^0.0.33",
-    "@prisma/client": "^5.13.0",
+    "@opentelemetry/api": ">=1.0.0 <1.9.0",
+    "@prisma/client": "^5.17.0",
     "@react-email/components": "^0.0.19",
     "@react-email/render": "^0.0.15",
     "@types/bcryptjs": "^2.4.6",
     "bcryptjs": "^2.4.3",
-    "kysely": "^0.27.3",
+    "bullmq": "^5.12.10",
+    "dd-trace": "^5.19.0",
+    "decimal.js": "^10.4.3",
+    "exponential-backoff": "^3.1.1",
+    "ioredis": "^5.4.1",
+    "kysely": "^0.27.4",
     "langchain": "^0.2.6",
     "lodash": "^4.17.21",
     "next-auth": "^4.24.7",
     "nodemailer": "^6.9.13",
     "prisma-extension-kysely": "^2.1.0",
+    "uuid": "^9.0.1",
+    "winston": "^3.14.2",
     "zod": "^3.23.8",
-    "zod-to-json-schema": "^3.22.5"
+    "zod-to-json-schema": "^3.23.2"
   },
   "devDependencies": {
-    "@repo/eslint-config": "*",
-    "@repo/typescript-config": "*",
-    "@types/lodash": "^4.17.5",
+    "@repo/eslint-config": "workspace:*",
+    "@repo/typescript-config": "workspace:*",
+    "@types/lodash": "^4.17.7",
     "@types/node": "^20.11.29",
     "@types/nodemailer": "^6.4.15",
     "@types/pg": "^8.11.6",
@@ -80,7 +99,7 @@
     "kysely-codegen": "^0.11.0",
     "nodemon": "^3.1.3",
     "prettier": "^3.2.5",
-    "prisma": "^5.13.0",
+    "prisma": "^5.17.0",
     "prisma-erd-generator": "^1.11.2",
     "prisma-kysely": "^1.8.0",
     "ts-node": "^10.9.2",
@@ -89,7 +108,7 @@
     "vitest": "^1.5.3"
   },
   "peerDependencies": {
-    "@types/react": "^18.2.79",
-    "react": "^18.0.0"
+    "@types/react": "~18.2.79",
+    "react": "~18.2.0"
   }
 }

packages/shared/prisma/generated/types.ts

@@ -4,13 +4,14 @@ export type Generated<T> = T extends ColumnType<infer S, infer I, infer U>
   : ColumnType<T, T | undefined, T>;
 export type Timestamp = ColumnType<Date, Date | string, Date | string>;
 
-export const ProjectRole = {
+export const Role = {
     OWNER: "OWNER",
     ADMIN: "ADMIN",
     MEMBER: "MEMBER",
-    VIEWER: "VIEWER"
+    VIEWER: "VIEWER",
+    NONE: "NONE"
 } as const;
-export type ProjectRole = (typeof ProjectRole)[keyof typeof ProjectRole];
+export type Role = (typeof Role)[keyof typeof Role];
 export const ObservationType = {
     SPAN: "SPAN",
     EVENT: "EVENT",
@@ -41,6 +42,13 @@ export const DatasetStatus = {
     ARCHIVED: "ARCHIVED"
 } as const;
 export type DatasetStatus = (typeof DatasetStatus)[keyof typeof DatasetStatus];
+export const CommentObjectType = {
+    TRACE: "TRACE",
+    OBSERVATION: "OBSERVATION",
+    SESSION: "SESSION",
+    PROMPT: "PROMPT"
+} as const;
+export type CommentObjectType = (typeof CommentObjectType)[keyof typeof CommentObjectType];
 export const JobType = {
     EVAL: "EVAL"
 } as const;
@@ -90,8 +98,10 @@ export type AuditLog = {
     created_at: Generated<Timestamp>;
     updated_at: Generated<Timestamp>;
     user_id: string;
-    project_id: string;
-    user_project_role: ProjectRole;
+    org_id: string;
+    user_org_role: string;
+    project_id: string | null;
+    user_project_role: string | null;
     resource_type: string;
     resource_id: string;
     action: string;
@@ -113,6 +123,16 @@ export type BatchExport = {
     url: string | null;
     log: string | null;
 };
+export type Comment = {
+    id: string;
+    project_id: string;
+    object_type: CommentObjectType;
+    object_id: string;
+    created_at: Generated<Timestamp>;
+    updated_at: Generated<Timestamp>;
+    content: string;
+    author_user_id: string | null;
+};
 export type CronJobs = {
     name: string;
     last_run: Timestamp | null;
@@ -121,15 +141,16 @@ export type CronJobs = {
 };
 export type Dataset = {
     id: string;
+    project_id: string;
     name: string;
     description: string | null;
     metadata: unknown | null;
-    project_id: string;
     created_at: Generated<Timestamp>;
     updated_at: Generated<Timestamp>;
 };
 export type DatasetItem = {
     id: string;
+    project_id: string;
     status: Generated<DatasetStatus>;
     input: unknown | null;
     expected_output: unknown | null;
@@ -142,6 +163,7 @@ export type DatasetItem = {
 };
 export type DatasetRunItems = {
     id: string;
+    project_id: string;
     dataset_run_id: string;
     dataset_item_id: string;
     trace_id: string;
@@ -151,6 +173,7 @@ export type DatasetRunItems = {
 };
 export type DatasetRuns = {
     id: string;
+    project_id: string;
     name: string;
     description: string | null;
     metadata: unknown | null;
@@ -226,9 +249,11 @@ export type LlmApiKeys = {
 export type MembershipInvitation = {
     id: string;
     email: string;
-    role: ProjectRole;
-    project_id: string;
-    sender_id: string | null;
+    org_id: string;
+    org_role: Role;
+    project_id: string | null;
+    project_role: Role | null;
+    invited_by_user_id: string | null;
     created_at: Generated<Timestamp>;
     updated_at: Generated<Timestamp>;
 };
@@ -306,6 +331,8 @@ export type ObservationView = {
     unit: string | null;
     completion_start_time: Timestamp | null;
     prompt_id: string | null;
+    prompt_name: string | null;
+    prompt_version: number | null;
     model_id: string | null;
     input_price: string | null;
     output_price: string | null;
@@ -316,6 +343,21 @@ export type ObservationView = {
     latency: number | null;
     time_to_first_token: number | null;
 };
+export type Organization = {
+    id: string;
+    name: string;
+    created_at: Generated<Timestamp>;
+    updated_at: Generated<Timestamp>;
+    cloud_config: unknown | null;
+};
+export type OrganizationMembership = {
+    id: string;
+    org_id: string;
+    user_id: string;
+    role: Role;
+    created_at: Generated<Timestamp>;
+    updated_at: Generated<Timestamp>;
+};
 export type PosthogIntegration = {
     project_id: string;
     encrypted_posthog_api_key: string;
@@ -326,15 +368,16 @@ export type PosthogIntegration = {
 };
 export type Project = {
     id: string;
+    org_id: string;
     created_at: Generated<Timestamp>;
     updated_at: Generated<Timestamp>;
     name: string;
-    cloud_config: unknown | null;
 };
 export type ProjectMembership = {
+    org_membership_id: string;
     project_id: string;
     user_id: string;
-    role: ProjectRole;
+    role: Role;
     created_at: Generated<Timestamp>;
     updated_at: Generated<Timestamp>;
 };
@@ -465,6 +508,7 @@ export type DB = {
     api_keys: ApiKey;
     audit_logs: AuditLog;
     batch_exports: BatchExport;
+    comments: Comment;
     cron_jobs: CronJobs;
     dataset_items: DatasetItem;
     dataset_run_items: DatasetRunItems;
@@ -479,6 +523,8 @@ export type DB = {
     models: Model;
     observations: Observation;
     observations_view: ObservationView;
+    organization_memberships: OrganizationMembership;
+    organizations: Organization;
     posthog_integrations: PosthogIntegration;
     project_memberships: ProjectMembership;
     projects: Project;

packages/shared/prisma/migrations/20240710114044_add_pricing_gpt4o_mini/migration.sql

@@ -0,0 +1,20 @@
+INSERT INTO models (
+  id,
+  project_id,
+  model_name,
+  match_pattern,
+  start_date,
+  input_price,
+  output_price,
+  total_price,
+  unit,
+  tokenizer_id,
+  tokenizer_config
+)
+VALUES
+  -- gpt-4o-mini
+  ('clyrjp56f0000t0mzapoocd7u', NULL, 'gpt-4o-mini', '(?i)^(gpt-4o-mini)$', NULL, 0.00000015, 0.0000006, NULL, 'TOKENS', 'openai', '{ "tokensPerMessage": 3, "tokensPerName": 1, "tokenizerModel": "gpt-4o" }'),
+  
+  -- gpt-4o-mini-2024-07-18
+  ('clyrjpbe20000t0mzcbwc42rg', NULL, 'gpt-4o-mini-2024-07-18', '(?i)^(gpt-4o-mini-2024-07-18)$', NULL, 0.00000015, 0.0000006, NULL, 'TOKENS', 'openai', '{ "tokensPerMessage": 3, "tokensPerName": 1, "tokenizerModel": "gpt-4o" }')
+  

packages/shared/prisma/migrations/20240718004923_datasets_tables_add_projectid_composite_key/migration.sql

@@ -0,0 +1,68 @@
+-- Add project_id to datasets, dataset_items, dataset_runs, dataset_run_items
+ALTER TABLE "dataset_items" 
+ADD COLUMN "project_id" TEXT;
+
+ALTER TABLE "dataset_run_items" 
+ADD COLUMN "project_id" TEXT;
+
+ALTER TABLE "dataset_runs" 
+ADD COLUMN "project_id" TEXT;
+
+-- Backfill project_id for dataset_items
+UPDATE dataset_items
+SET project_id = datasets.project_id
+FROM datasets
+WHERE dataset_items.dataset_id = datasets.id;
+
+-- Backfill project_id for dataset_runs
+UPDATE dataset_runs
+SET project_id = datasets.project_id
+FROM datasets
+WHERE dataset_runs.dataset_id = datasets.id;
+
+-- Backfill project_id for dataset_run_items
+UPDATE dataset_run_items
+SET project_id = dataset_runs.project_id
+FROM dataset_runs
+WHERE dataset_run_items.dataset_run_id = dataset_runs.id;
+
+-- Drop the old foreign keys
+ALTER TABLE "dataset_run_items" 
+DROP CONSTRAINT "dataset_run_items_dataset_item_id_fkey",
+DROP CONSTRAINT "dataset_run_items_dataset_run_id_fkey";
+
+ALTER TABLE "dataset_items" 
+DROP CONSTRAINT "dataset_items_dataset_id_fkey";
+
+ALTER TABLE "dataset_runs" 
+DROP CONSTRAINT "dataset_runs_dataset_id_fkey";
+
+-- Now alter the columns to NOT NULL and update primary keys
+ALTER TABLE "datasets" 
+DROP CONSTRAINT "datasets_pkey",
+ADD CONSTRAINT "datasets_pkey" PRIMARY KEY ("id", "project_id");
+
+ALTER TABLE "dataset_items" 
+ALTER COLUMN "project_id" SET NOT NULL,
+DROP CONSTRAINT "dataset_items_pkey",
+ADD CONSTRAINT "dataset_items_pkey" PRIMARY KEY ("id", "project_id");
+
+ALTER TABLE "dataset_runs" 
+ALTER COLUMN "project_id" SET NOT NULL,
+DROP CONSTRAINT "dataset_runs_pkey",
+ADD CONSTRAINT "dataset_runs_pkey" PRIMARY KEY ("id", "project_id");
+
+ALTER TABLE "dataset_run_items" 
+ALTER COLUMN "project_id" SET NOT NULL,
+DROP CONSTRAINT "dataset_run_items_pkey",
+ADD CONSTRAINT "dataset_run_items_pkey" PRIMARY KEY ("id", "project_id");
+
+
+-- Add new foreign keys
+ALTER TABLE "dataset_items" ADD CONSTRAINT "dataset_items_dataset_id_project_id_fkey" FOREIGN KEY ("dataset_id", "project_id") REFERENCES "datasets"("id", "project_id") ON DELETE CASCADE ON UPDATE CASCADE;
+
+ALTER TABLE "dataset_runs" ADD CONSTRAINT "dataset_runs_dataset_id_project_id_fkey" FOREIGN KEY ("dataset_id", "project_id") REFERENCES "datasets"("id", "project_id") ON DELETE CASCADE ON UPDATE CASCADE;
+
+ALTER TABLE "dataset_run_items" ADD CONSTRAINT "dataset_run_items_dataset_run_id_project_id_fkey" FOREIGN KEY ("dataset_run_id", "project_id") REFERENCES "dataset_runs"("id", "project_id") ON DELETE CASCADE ON UPDATE CASCADE;
+
+ALTER TABLE "dataset_run_items" ADD CONSTRAINT "dataset_run_items_dataset_item_id_project_id_fkey" FOREIGN KEY ("dataset_item_id", "project_id") REFERENCES "dataset_items"("id", "project_id") ON DELETE CASCADE ON UPDATE CASCADE;

packages/shared/prisma/migrations/20240718011733_dataset_runs_add_unique_dataset_id_project_id_name copy/migration.sql

@@ -0,0 +1,2 @@
+-- CreateIndex
+CREATE UNIQUE INDEX CONCURRENTLY "dataset_runs_dataset_id_project_id_name_key" ON "dataset_runs"("dataset_id", "project_id", "name");

packages/shared/prisma/migrations/20240718011734_dataset_runs_drop_unique_dataset_id_name/migration.sql

@@ -0,0 +1,2 @@
+-- DropIndex
+DROP INDEX CONCURRENTLY "dataset_runs_dataset_id_name_key";

packages/shared/prisma/migrations/20240718011735_observation_view_add_prompt_name_and_version/migration.sql

@@ -0,0 +1,100 @@
+DROP VIEW IF EXISTS "observations_view"; -- Drop view as column was added in 20240705154048_observation_view_add_created_at_updated_at and update view must have same columns
+CREATE VIEW "observations_view" AS -- Specify the columns that should be returned in the view, as calculated columns are added but exist in the observations table already
+SELECT
+    o.id,
+    o.name,
+    o.start_time,
+    o.end_time,
+    o.parent_observation_id,
+    o.type,
+    o.trace_id,
+    o.metadata,
+    o.model,
+    o."modelParameters",
+    o.input,
+    o.output,
+    o.level,
+    o.status_message,
+    o.completion_start_time,
+    o.completion_tokens,
+    o.prompt_tokens,
+    o.total_tokens,
+    o.version,
+    o.project_id,
+    o.created_at,
+    o.updated_at,
+    o.unit,
+    o.prompt_id,
+    p.name as prompt_name,         -- added in this change
+    p.version as prompt_version,   -- added in this change
+    o.input_cost,
+    o.output_cost,
+    o.total_cost,
+    o.internal_model,
+    m.id AS "model_id",
+    m.start_date AS "model_start_date",
+    m.input_price,
+    m.output_price,
+    m.total_price,
+    m.tokenizer_config AS "tokenizer_config",
+    CASE 
+        WHEN o.calculated_input_cost IS NULL AND o.input_cost IS NULL AND o.output_cost IS NULL AND o.total_cost IS NULL THEN
+            o.prompt_tokens::decimal * m.input_price
+        ELSE
+            COALESCE(o.calculated_input_cost, o.input_cost)
+    END AS "calculated_input_cost",
+    CASE 
+        WHEN o.calculated_output_cost IS NULL AND o.input_cost IS NULL AND o.output_cost IS NULL AND o.total_cost IS NULL THEN
+            o.completion_tokens::decimal * m.output_price
+        ELSE
+            COALESCE(o.calculated_output_cost, o.output_cost)
+    END AS "calculated_output_cost",
+    CASE 
+        WHEN o.calculated_total_cost IS NULL AND o.input_cost IS NULL AND o.output_cost IS NULL AND o.total_cost IS NULL THEN
+            CASE 
+                WHEN m.total_price IS NOT NULL AND o.total_tokens IS NOT NULL THEN
+                    m.total_price * o.total_tokens
+                ELSE
+                    o.prompt_tokens::decimal * m.input_price + 
+                    o.completion_tokens::decimal * m.output_price
+            END
+        ELSE
+            COALESCE(o.calculated_total_cost, o.total_cost)
+    END AS "calculated_total_cost",
+    CASE WHEN o.end_time IS NULL THEN NULL ELSE (EXTRACT(EPOCH FROM o."end_time") - EXTRACT(EPOCH FROM o."start_time"))::double precision END AS "latency",
+    CASE WHEN o.completion_start_time IS NOT NULL AND o.start_time IS NOT NULL THEN EXTRACT(EPOCH FROM (completion_start_time - start_time))::double precision ELSE NULL END as "time_to_first_token"
+    
+FROM
+    observations o
+LEFT JOIN LATERAL (
+    SELECT
+        models.*
+    FROM
+        models
+    WHERE (models.project_id = o.project_id OR models.project_id IS NULL)
+    AND models.model_name = o.internal_model
+    AND (models.start_date < o.start_time OR models.start_date IS NULL)
+    AND o.unit::TEXT = models.unit
+    ORDER BY
+        models.project_id ASC, -- in postgres, NULLs are sorted last when ordering ASC
+        models.start_date DESC NULLS LAST -- now, NULLs are sorted last when ordering DESC as well
+    LIMIT 1
+) m ON TRUE
+LEFT JOIN LATERAL (
+    SELECT
+        prompts.*
+    FROM
+        prompts
+    WHERE prompts.id = o.prompt_id
+    AND prompts.project_id = o.project_id
+    LIMIT 1
+) p ON TRUE
+
+
+-- requirements:
+-- 1. The view should return all columns from the observations table
+-- 2. The view should match with only one model for each observation if:
+--     a. The model has the same project_id as the observation, otherwise the model without project_id. 
+--     b. The model has the same model_name as the observation
+--     c. The model has a start_date that is less than the observation start_time, otherwise the model without start_date
+--     d. The model has the same unit as the observation

packages/shared/prisma/migrations/20240807111358_models_add_openai_gpt_4o_2024_08_06/migration.sql

@@ -0,0 +1,18 @@
+
+INSERT INTO models (
+  id,
+  project_id,
+  model_name,
+  match_pattern,
+  start_date,
+  input_price,
+  output_price,
+  total_price,
+  unit,
+  tokenizer_id,
+  tokenizer_config
+)
+VALUES
+  -- gpt-4o-2024-08-06
+  ('clzjr85f70000ymmzg7hqffra', NULL, 'gpt-4o-2024-08-06', '(?i)^(gpt-4o-2024-08-06)$', NULL, 0.0000025, 0.000010, NULL, 'TOKENS', 'openai', '{ "tokensPerMessage": 3, "tokensPerName": 1, "tokenizerModel": "gpt-4o" }')
+  

packages/shared/prisma/migrations/20240807111359_add_organizations_main_migration/migration.sql

@@ -0,0 +1,164 @@
+-- Table ORGANIZATIONS
+-- Create empty table ORGANIZATIONS
+CREATE TABLE "organizations" (
+    "id" TEXT NOT NULL,
+    "name" TEXT NOT NULL,
+    "created_at" TIMESTAMP(3) NOT NULL DEFAULT CURRENT_TIMESTAMP,
+    "updated_at" TIMESTAMP(3) NOT NULL DEFAULT CURRENT_TIMESTAMP,
+    "cloud_config" JSONB,
+
+    CONSTRAINT "organizations_pkey" PRIMARY KEY ("id")
+);
+-- Backfill table ORGANIZATIONS: create a new organization for each project, move cloudConfig, and set the org_id on the project to project_id prefixed with 'o'
+INSERT INTO "organizations" ("id", "name", "cloud_config", "created_at")
+SELECT
+  CONCAT('o', "id") as "id", -- This mapping is used in other migration steps as well, keep it consistent
+  "name",
+  "cloud_config",
+  "created_at"
+FROM "projects";
+-- Drop column cloud_config from projects as it's now on organization level
+ALTER TABLE "projects" DROP COLUMN "cloud_config";
+
+
+-- PROJECT.ORG_ID
+-- Add org_id to projects
+ALTER TABLE "projects" ADD COLUMN "org_id" TEXT;
+-- Backfill: org_id on projects, set it to project_id prefixed with 'o'
+UPDATE "projects"
+SET "org_id" = CONCAT('o', "id");
+-- Set not null after backfill
+ALTER TABLE "projects" ALTER COLUMN "org_id" SET NOT NULL;
+-- AddForeignKey
+ALTER TABLE "projects" ADD CONSTRAINT "projects_org_id_fkey" FOREIGN KEY ("org_id") REFERENCES "organizations"("id") ON DELETE CASCADE ON UPDATE CASCADE;
+-- CreateIndex
+CREATE INDEX "projects_org_id_idx" ON "projects"("org_id");
+
+
+-- ORGANIZATION MEMBERSHIPS
+-- Create UserRole ENUM
+CREATE TYPE "Role" AS ENUM ('OWNER', 'ADMIN', 'MEMBER', 'VIEWER', 'NONE');
+-- Create empty table
+CREATE TABLE "organization_memberships" (
+    "id" TEXT NOT NULL,
+    "org_id" TEXT NOT NULL,
+    "user_id" TEXT NOT NULL,
+    "role" "Role" NOT NULL,
+    "created_at" TIMESTAMP(3) NOT NULL DEFAULT CURRENT_TIMESTAMP,
+    "updated_at" TIMESTAMP(3) NOT NULL DEFAULT CURRENT_TIMESTAMP,
+
+    CONSTRAINT "organization_memberships_pkey" PRIMARY KEY ("id")
+);
+-- Add indexes
+CREATE INDEX "organization_memberships_user_id_idx" ON "organization_memberships"("user_id");
+CREATE INDEX "organization_memberships_org_id_idx" ON "organization_memberships"("org_id");
+CREATE UNIQUE INDEX "organization_memberships_org_id_user_id_key" ON "organization_memberships"("org_id", "user_id");
+-- AddForeignKey
+ALTER TABLE "organization_memberships" ADD CONSTRAINT "organization_memberships_org_id_fkey" FOREIGN KEY ("org_id") REFERENCES "organizations"("id") ON DELETE CASCADE ON UPDATE CASCADE;
+-- AddForeignKey
+ALTER TABLE "organization_memberships" ADD CONSTRAINT "organization_memberships_user_id_fkey" FOREIGN KEY ("user_id") REFERENCES "users"("id") ON DELETE CASCADE ON UPDATE CASCADE;
+
+
+
+-- Migrate project memberships to organization memberships
+INSERT INTO "organization_memberships" ("id", "org_id", "user_id", "role", "created_at", "updated_at")
+SELECT
+  md5(random()::text || clock_timestamp()::text || project_id::text || user_id::text)::uuid AS "id",
+  CONCAT('o', "project_id") as "org_id",
+  "user_id",
+  "role"::text::"Role" as "role",
+  "created_at",
+  "updated_at"
+FROM "project_memberships";
+
+
+
+-- Delete all project memberships after migration to organization memberships
+DELETE FROM "project_memberships";
+
+-- Add org_membership_id to project_memberships, ok to be not null as it's a new column on a now empty table
+ALTER TABLE "project_memberships" ADD COLUMN     "org_membership_id" TEXT NOT NULL;
+-- Switch to new UserRole enum
+ALTER TABLE "project_memberships" DROP COLUMN "role",
+ADD COLUMN     "role" "Role" NOT NULL;
+-- AddForeignKey
+ALTER TABLE "project_memberships" ADD CONSTRAINT "project_memberships_org_membership_id_fkey" FOREIGN KEY ("org_membership_id") REFERENCES "organization_memberships"("id") ON DELETE CASCADE ON UPDATE CASCADE;
+-- CreateIndex
+CREATE INDEX "project_memberships_project_id_idx" ON "project_memberships"("project_id");
+-- CreateIndex
+CREATE INDEX "project_memberships_org_membership_id_idx" ON "project_memberships"("org_membership_id");
+
+
+
+-- AUDIT LOGS
+-- DropForeignKey, just index on these, should remain after project/org/user deletions
+ALTER TABLE "audit_logs" DROP CONSTRAINT "audit_logs_user_id_fkey";
+ALTER TABLE "audit_logs" DROP CONSTRAINT "audit_logs_project_id_fkey";
+-- Drop not null on project level cols
+ALTER TABLE "audit_logs" ALTER COLUMN "project_id" DROP NOT NULL;
+ALTER TABLE "audit_logs" ALTER COLUMN "user_project_role" DROP NOT NULL;
+-- Add org-level cols
+ALTER TABLE "audit_logs" ADD COLUMN "org_id" TEXT;
+ALTER TABLE "audit_logs" ADD COLUMN "user_org_role" TEXT;
+-- Backfill org_id
+UPDATE "audit_logs"
+SET "org_id" = "projects"."org_id"
+FROM "projects"
+WHERE "audit_logs"."project_id" = "projects"."id";
+-- Backfill user_org_role with value from user_project_role
+UPDATE "audit_logs"
+SET "user_org_role" = "user_project_role"::text::"Role";
+-- Drop and recreate user_project_role column as text column going forward, empty for historical data as it's all on org level now
+ALTER TABLE "audit_logs" DROP COLUMN "user_project_role";
+ALTER TABLE "audit_logs" ADD COLUMN "user_project_role" TEXT; -- nullable
+-- Add not null on org level cols
+ALTER TABLE "audit_logs" ALTER COLUMN "org_id" SET NOT NULL;
+ALTER TABLE "audit_logs" ALTER COLUMN "user_org_role" SET NOT NULL;
+-- CreateIndex
+CREATE INDEX "audit_logs_user_id_idx" ON "audit_logs"("user_id");
+CREATE INDEX "audit_logs_org_id_idx" ON "audit_logs"("org_id");
+
+
+-- MEMBERSHIP INVITATIONS
+-- Rename sender_id to invited_by_user_id
+ALTER TABLE "membership_invitations" DROP CONSTRAINT "membership_invitations_sender_id_fkey";
+ALTER TABLE "membership_invitations" RENAME COLUMN "sender_id" TO "invited_by_user_id";
+ALTER TABLE "membership_invitations" ADD CONSTRAINT "membership_invitations_invited_by_user_id_fkey" FOREIGN KEY ("invited_by_user_id") REFERENCES "users"("id") ON DELETE SET NULL ON UPDATE CASCADE;
+-- DropForeignKey
+ALTER TABLE "membership_invitations" DROP CONSTRAINT "membership_invitations_project_id_fkey";
+-- AlterTable
+ALTER TABLE "membership_invitations" RENAME COLUMN "role" TO "project_role";
+ALTER TABLE "membership_invitations" ADD COLUMN "org_id" TEXT;
+ALTER TABLE "membership_invitations" ADD COLUMN "org_role" "Role";
+ALTER TABLE "membership_invitations" ALTER COLUMN "project_id" DROP NOT NULL;
+ALTER TABLE "membership_invitations" ALTER COLUMN "project_role" DROP NOT NULL;
+-- Backfill org id
+UPDATE "membership_invitations"
+SET "org_id" = "projects"."org_id"
+FROM "projects"
+WHERE "membership_invitations"."project_id" = "projects"."id";
+-- AddForeignKey
+ALTER TABLE "membership_invitations" ADD CONSTRAINT "membership_invitations_org_id_fkey" FOREIGN KEY ("org_id") REFERENCES "organizations"("id") ON DELETE CASCADE ON UPDATE CASCADE;
+-- Backfill org role with value from project role
+UPDATE "membership_invitations"
+SET "org_role" = "project_role"::text::"Role";
+-- Set project-level cols to null, as it's now org level for all existing invitations and role enum will change below
+UPDATE "membership_invitations"
+SET "project_role" = NULL, "project_id" = NULL;
+-- Switch to new UserRole enum
+ALTER TABLE "membership_invitations" DROP COLUMN "project_role",
+ADD COLUMN     "project_role" "Role";
+
+-- AddForeignKey
+ALTER TABLE "membership_invitations" ADD CONSTRAINT "membership_invitations_project_id_fkey" FOREIGN KEY ("project_id") REFERENCES "projects"("id") ON DELETE SET NULL ON UPDATE CASCADE;
+
+-- Add not null on org level cols after backfill
+ALTER TABLE "membership_invitations"
+ALTER COLUMN "org_id" SET NOT NULL,
+ALTER COLUMN "org_role" SET NOT NULL;
+-- CreateIndex
+CREATE INDEX "membership_invitations_org_id_idx" ON "membership_invitations"("org_id");
+
+
+-- Drop ProjectRole enum as it is replaced by Role
+DROP TYPE "ProjectRole";

packages/shared/prisma/migrations/20240814223824_model_fix_text_embedding_3_large/migration.sql

@@ -0,0 +1,7 @@
+
+-- Fix model name
+-- https://github.com/langfuse/langfuse/issues/2688
+
+UPDATE models
+SET model_name = 'text-embedding-3-large'
+WHERE id = 'clruwn76700020al7gp8e4g4l'

packages/shared/prisma/migrations/20240814233029_dataset_items_drop_fkey_on_traces_and_observations/migration.sql

@@ -0,0 +1,5 @@
+-- DropForeignKey
+ALTER TABLE "dataset_items" DROP CONSTRAINT "dataset_items_source_observation_id_fkey";
+
+-- DropForeignKey
+ALTER TABLE "dataset_items" DROP CONSTRAINT "dataset_items_source_trace_id_fkey";

packages/shared/prisma/migrations/20240815171916_add_comments/migration.sql

@@ -0,0 +1,22 @@
+-- CreateEnum
+CREATE TYPE "CommentObjectType" AS ENUM ('TRACE', 'OBSERVATION', 'SESSION', 'PROMPT');
+
+-- CreateTable
+CREATE TABLE "comments" (
+    "id" TEXT NOT NULL,
+    "project_id" TEXT NOT NULL,
+    "object_type" "CommentObjectType" NOT NULL,
+    "object_id" TEXT NOT NULL,
+    "created_at" TIMESTAMP(3) NOT NULL DEFAULT CURRENT_TIMESTAMP,
+    "updated_at" TIMESTAMP(3) NOT NULL DEFAULT CURRENT_TIMESTAMP,
+    "content" TEXT NOT NULL,
+    "author_user_id" TEXT,
+
+    CONSTRAINT "comments_pkey" PRIMARY KEY ("id")
+);
+
+-- CreateIndex
+CREATE INDEX "comments_project_id_object_type_object_id_idx" ON "comments"("project_id", "object_type", "object_id");
+
+-- AddForeignKey
+ALTER TABLE "comments" ADD CONSTRAINT "comments_project_id_fkey" FOREIGN KEY ("project_id") REFERENCES "projects"("id") ON DELETE CASCADE ON UPDATE CASCADE;

packages/shared/prisma/schema.prisma

@@ -3,7 +3,7 @@
 
 generator client {
     provider        = "prisma-client-js"
-    previewFeatures = ["tracing", "views", "relationJoins"]
+    previewFeatures = ["tracing", "views", "relationJoins", "metrics"]
 }
 
 datasource db {
@@ -63,21 +63,21 @@ model Session {
 }
 
 model User {
-    id                 String                 @id @default(cuid())
-    name               String?
-    email              String?                @unique
-    emailVerified      DateTime?              @map("email_verified")
-    password           String?
-    image              String?
-    admin              Boolean                @default(false)
-    accounts           Account[]
-    sessions           Session[]
-    projectMemberships ProjectMembership[]
-    invitations        MembershipInvitation[]
-    createdAt          DateTime               @default(now()) @map("created_at")
-    updatedAt          DateTime               @default(now()) @updatedAt @map("updated_at")
-    featureFlags       String[]               @default([]) @map("feature_flags")
-    AuditLog           AuditLog[]
+    id                      String                   @id @default(cuid())
+    name                    String?
+    email                   String?                  @unique
+    emailVerified           DateTime?                @map("email_verified")
+    password                String?
+    image                   String?
+    admin                   Boolean                  @default(false)
+    accounts                Account[]
+    sessions                Session[]
+    organizationMemberships OrganizationMembership[]
+    projectMemberships      ProjectMembership[]
+    invitations             MembershipInvitation[]
+    createdAt               DateTime                 @default(now()) @map("created_at")
+    updatedAt               DateTime                 @default(now()) @updatedAt @map("updated_at")
+    featureFlags            String[]                 @default([]) @map("feature_flags")
 
     @@map("users")
 }
@@ -91,13 +91,27 @@ model VerificationToken {
     @@map("verification_tokens")
 }
 
+model Organization {
+    id                      String                   @id @default(cuid())
+    name                    String
+    createdAt               DateTime                 @default(now()) @map("created_at")
+    updatedAt               DateTime                 @default(now()) @updatedAt @map("updated_at")
+    cloudConfig             Json?                    @map("cloud_config") // Langfuse Cloud, for zod schema see @/src/features/organizations/utils/cloudConfigSchema
+    organizationMemberships OrganizationMembership[]
+    projects                Project[]
+    MembershipInvitation    MembershipInvitation[]
+
+    @@map("organizations")
+}
+
 model Project {
     id                 String                 @id @default(cuid())
+    orgId              String                 @map("org_id")
     createdAt          DateTime               @default(now()) @map("created_at")
     updatedAt          DateTime               @default(now()) @updatedAt @map("updated_at")
     name               String
-    cloudConfig        Json?                  @map("cloud_config") // Langfuse Cloud, for zod schema see projectsRouter.ts
     projectMembers     ProjectMembership[]
+    organization       Organization           @relation(fields: [orgId], references: [id], onUpdate: Cascade, onDelete: Cascade)
     traces             Trace[]
     observations       Observation[]
     apiKeys            ApiKey[]
@@ -107,7 +121,6 @@ model Project {
     sessions           TraceSession[]
     Prompt             Prompt[]
     Model              Model[]
-    AuditLog           AuditLog[]
     EvalTemplate       EvalTemplate[]
     JobConfiguration   JobConfiguration[]
     JobExecution       JobExecution[]
@@ -116,7 +129,9 @@ model Project {
     Score              Score[]
     scoreConfig        ScoreConfig[]
     BatchExport        BatchExport[]
+    comment            Comment[]
 
+    @@index([orgId])
     @@map("projects")
 }
 
@@ -161,41 +176,68 @@ model LlmApiKeys {
     @@map("llm_api_keys")
 }
 
+model OrganizationMembership {
+    id                 String              @id @default(cuid())
+    orgId              String              @map("org_id")
+    organization       Organization        @relation(fields: [orgId], references: [id], onDelete: Cascade)
+    userId             String              @map("user_id")
+    user               User                @relation(fields: [userId], references: [id], onDelete: Cascade)
+    role               Role                @map("role")
+    createdAt          DateTime            @default(now()) @map("created_at")
+    updatedAt          DateTime            @default(now()) @updatedAt @map("updated_at")
+    ProjectMemberships ProjectMembership[]
+
+    @@unique([orgId, userId])
+    @@index([userId])
+    @@index([orgId])
+    @@map("organization_memberships")
+}
+
+// Set a project-specific role for a user in an organization
 model ProjectMembership {
-    projectId String      @map("project_id")
-    project   Project     @relation(fields: [projectId], references: [id], onDelete: Cascade)
-    userId    String      @map("user_id")
-    user      User        @relation(fields: [userId], references: [id], onDelete: Cascade)
-    role      ProjectRole
-    createdAt DateTime    @default(now()) @map("created_at")
-    updatedAt DateTime    @default(now()) @updatedAt @map("updated_at")
+    orgMembershipId        String                 @map("org_membership_id")
+    organizationMembership OrganizationMembership @relation(fields: [orgMembershipId], references: [id], onDelete: Cascade)
+    projectId              String                 @map("project_id")
+    project                Project                @relation(fields: [projectId], references: [id], onDelete: Cascade)
+    userId                 String                 @map("user_id")
+    user                   User                   @relation(fields: [userId], references: [id], onDelete: Cascade)
+    role                   Role
+    createdAt              DateTime               @default(now()) @map("created_at")
+    updatedAt              DateTime               @default(now()) @updatedAt @map("updated_at")
 
     @@id([projectId, userId])
     @@index([userId])
+    @@index([projectId])
+    @@index([orgMembershipId])
     @@map("project_memberships")
 }
 
 model MembershipInvitation {
-    id        String      @id @unique @default(cuid())
-    email     String
-    role      ProjectRole
-    projectId String      @map("project_id")
-    project   Project     @relation(fields: [projectId], references: [id], onDelete: Cascade)
-    senderId  String?     @map("sender_id")
-    sender    User?       @relation(fields: [senderId], references: [id], onDelete: SetNull)
-    createdAt DateTime    @default(now()) @map("created_at")
-    updatedAt DateTime    @default(now()) @updatedAt @map("updated_at")
+    id              String       @id @unique @default(cuid())
+    email           String
+    orgId           String       @map("org_id")
+    organization    Organization @relation(fields: [orgId], references: [id], onDelete: Cascade)
+    orgRole         Role         @map("org_role")
+    projectId       String?      @map("project_id")
+    project         Project?     @relation(fields: [projectId], references: [id], onDelete: SetNull)
+    projectRole     Role?        @map("project_role")
+    invitedByUserId String?      @map("invited_by_user_id")
+    invitedByUser   User?        @relation(fields: [invitedByUserId], references: [id], onDelete: SetNull)
+    createdAt       DateTime     @default(now()) @map("created_at")
+    updatedAt       DateTime     @default(now()) @updatedAt @map("updated_at")
 
     @@index([projectId])
+    @@index([orgId])
     @@index([email])
     @@map("membership_invitations")
 }
 
-enum ProjectRole {
+enum Role {
     OWNER
     ADMIN
     MEMBER
     VIEWER
+    NONE
 }
 
 model TraceSession {
@@ -238,7 +280,6 @@ model Trace {
     createdAt  DateTime      @default(now()) @map("created_at")
     updatedAt  DateTime      @default(now()) @updatedAt @map("updated_at")
 
-    DatasetItem  DatasetItem[]
     JobExecution JobExecution[]
 
     @@index([projectId])
@@ -323,9 +364,8 @@ model Observation {
     calculatedOutputCost Decimal? @map("calculated_output_cost")
     calculatedTotalCost  Decimal? @map("calculated_total_cost")
 
-    completionStartTime DateTime?     @map("completion_start_time")
-    project             Project       @relation(fields: [projectId], references: [id], onDelete: Cascade)
-    derivedDatasetItems DatasetItem[]
+    completionStartTime DateTime? @map("completion_start_time")
+    project             Project   @relation(fields: [projectId], references: [id], onDelete: Cascade)
 
     promptId String? @map("prompt_id") // no fk constraint, prompt can be deleted
 
@@ -376,7 +416,10 @@ view ObservationView {
     unit                String?
     completionStartTime DateTime? @map("completion_start_time")
 
-    promptId String? @map("prompt_id")
+    // prompts
+    promptId      String? @map("prompt_id")
+    promptName    String? @map("prompt_name")
+    promptVersion Int?    @map("prompt_version")
 
     // model fields
     modelId     String?  @map("model_id")
@@ -488,17 +531,18 @@ model CronJobs {
 }
 
 model Dataset {
-    id           String        @id @default(cuid())
+    id           String        @default(cuid())
+    projectId    String        @map("project_id")
     name         String
     description  String?
     metadata     Json?
-    projectId    String        @map("project_id")
     project      Project       @relation(fields: [projectId], references: [id], onDelete: Cascade)
     createdAt    DateTime      @default(now()) @map("created_at")
     updatedAt    DateTime      @default(now()) @updatedAt @map("updated_at")
     datasetItems DatasetItem[]
     datasetRuns  DatasetRuns[]
 
+    @@id([id, projectId])
     @@unique([projectId, name])
     @@index([projectId], type: Hash)
     @@index([createdAt])
@@ -507,21 +551,21 @@ model Dataset {
 }
 
 model DatasetItem {
-    id                  String            @id @default(cuid())
+    id                  String            @default(cuid())
+    projectId           String            @map("project_id")
     status              DatasetStatus     @default(ACTIVE)
     input               Json?
     expectedOutput      Json?             @map("expected_output")
     metadata            Json?
     sourceTraceId       String?           @map("source_trace_id")
-    sourceTrace         Trace?            @relation(fields: [sourceTraceId], references: [id], onDelete: SetNull)
     sourceObservationId String?           @map("source_observation_id")
-    sourceObservation   Observation?      @relation(fields: [sourceObservationId], references: [id], onDelete: SetNull)
     datasetId           String            @map("dataset_id")
-    dataset             Dataset           @relation(fields: [datasetId], references: [id], onDelete: Cascade)
+    dataset             Dataset           @relation(fields: [datasetId, projectId], references: [id, projectId], onDelete: Cascade)
     createdAt           DateTime          @default(now()) @map("created_at")
     updatedAt           DateTime          @default(now()) @updatedAt @map("updated_at")
     datasetRunItems     DatasetRunItems[]
 
+    @@id([id, projectId])
     @@index([sourceTraceId], type: Hash)
     @@index([sourceObservationId], type: Hash)
     @@index([datasetId], type: Hash)
@@ -536,17 +580,19 @@ enum DatasetStatus {
 }
 
 model DatasetRuns {
-    id              String            @id @default(cuid())
+    id              String            @default(cuid())
+    projectId       String            @map("project_id")
     name            String
     description     String?
     metadata        Json?
     datasetId       String            @map("dataset_id")
-    dataset         Dataset           @relation(fields: [datasetId], references: [id], onDelete: Cascade)
+    dataset         Dataset           @relation(fields: [datasetId, projectId], references: [id, projectId], onDelete: Cascade)
     createdAt       DateTime          @default(now()) @map("created_at")
     updatedAt       DateTime          @default(now()) @updatedAt @map("updated_at")
     datasetRunItems DatasetRunItems[]
 
-    @@unique([datasetId, name])
+    @@id([id, projectId])
+    @@unique([datasetId, projectId, name])
     @@index([datasetId], type: Hash)
     @@index([createdAt])
     @@index([updatedAt])
@@ -554,16 +600,18 @@ model DatasetRuns {
 }
 
 model DatasetRunItems {
-    id            String      @id @default(cuid())
+    id            String      @default(cuid())
+    projectId     String      @map("project_id")
     datasetRunId  String      @map("dataset_run_id")
-    datasetRun    DatasetRuns @relation(fields: [datasetRunId], references: [id], onDelete: Cascade)
+    datasetRun    DatasetRuns @relation(fields: [datasetRunId, projectId], references: [id, projectId], onDelete: Cascade)
     datasetItemId String      @map("dataset_item_id")
-    datasetItem   DatasetItem @relation(fields: [datasetItemId], references: [id], onDelete: Cascade)
+    datasetItem   DatasetItem @relation(fields: [datasetItemId, projectId], references: [id, projectId], onDelete: Cascade)
     traceId       String      @map("trace_id")
     observationId String?     @map("observation_id")
     createdAt     DateTime    @default(now()) @map("created_at")
     updatedAt     DateTime    @default(now()) @updatedAt @map("updated_at")
 
+    @@id([id, projectId])
     @@index([datasetRunId], type: Hash)
     @@index([datasetItemId], type: Hash)
     @@index([observationId], type: Hash)
@@ -588,6 +636,28 @@ model Events {
     @@map("events")
 }
 
+model Comment {
+    id           String            @id @default(cuid())
+    projectId    String            @map("project_id")
+    project      Project           @relation(fields: [projectId], references: [id], onDelete: Cascade)
+    objectType   CommentObjectType @map("object_type")
+    objectId     String            @map("object_id")
+    createdAt    DateTime          @default(now()) @map("created_at")
+    updatedAt    DateTime          @default(now()) @updatedAt @map("updated_at")
+    content      String
+    authorUserId String?           @map("author_user_id") // no fk constraint, user can be deleted
+
+    @@index([projectId, objectType, objectId])
+    @@map("comments")
+}
+
+enum CommentObjectType {
+    TRACE
+    OBSERVATION
+    SESSION
+    PROMPT
+}
+
 model Prompt {
     id        String   @id @default(cuid())
     createdAt DateTime @default(now()) @map("created_at")
@@ -632,7 +702,7 @@ model Model {
     inputPrice      Decimal?  @map("input_price")
     outputPrice     Decimal?  @map("output_price")
     totalPrice      Decimal?  @map("total_price")
-    unit            String // TOKENS, CHARACTERS, MILLISECONDS, SECONDS, or IMAGES
+    unit            String // TOKENS, CHARACTERS, MILLISECONDS, SECONDS, REQUESTS, or IMAGES
     tokenizerId     String?   @map("tokenizer_id")
     tokenizerConfig Json?     @map("tokenizer_config")
 
@@ -643,22 +713,25 @@ model Model {
     @@map("models")
 }
 
+// No FK constraints to preserve audit logs
 model AuditLog {
-    id              String      @id @default(cuid())
-    createdAt       DateTime    @default(now()) @map("created_at")
-    updatedAt       DateTime    @default(now()) @updatedAt @map("updated_at")
-    userId          String      @map("user_id")
-    user            User        @relation(fields: [userId], references: [id], onDelete: Cascade)
-    projectId       String      @map("project_id")
-    project         Project     @relation(fields: [projectId], references: [id], onDelete: Cascade)
-    userProjectRole ProjectRole @map("user_project_role")
-    resourceType    String      @map("resource_type")
-    resourceId      String      @map("resource_id")
+    id              String   @id @default(cuid())
+    createdAt       DateTime @default(now()) @map("created_at")
+    updatedAt       DateTime @default(now()) @updatedAt @map("updated_at")
+    userId          String   @map("user_id")
+    orgId           String   @map("org_id")
+    userOrgRole     String   @map("user_org_role")
+    projectId       String?  @map("project_id")
+    userProjectRole String?  @map("user_project_role")
+    resourceType    String   @map("resource_type")
+    resourceId      String   @map("resource_id")
     action          String
     before          String? //stringified JSON
     after           String? // stringified JSON
 
     @@index([projectId])
+    @@index([userId])
+    @@index([orgId])
     @@index([createdAt])
     @@index([updatedAt])
     @@map("audit_logs")

packages/shared/prisma/seed.ts

@@ -12,8 +12,9 @@ import { parseArgs } from "node:util";
 import { chunk } from "lodash";
 import { v4 } from "uuid";
 import { ModelUsageUnit } from "../src";
-import { getDisplaySecretKey, hashSecretKey } from "../src/server";
+import { getDisplaySecretKey, hashSecretKey, logger } from "../src/server";
 import { encrypt } from "../src/encryption";
+import { redis } from "../src/server/redis/redis";
 
 const LOAD_TRACE_VOLUME = 10_000;
 
@@ -33,39 +34,111 @@ async function main() {
     options,
   }).values.environment;
 
+  const seedOrgId = "seed-org-id";
+  const seedProjectId = "7a88fb47-b4e2-43b8-a06c-a5ce950dc53a";
+  const seedUserId1 = "user-1"; // Owner of org
+  const seedUserId2 = "user-2"; // Member of org, admin of project
+
   const user = await prisma.user.upsert({
-    where: { id: "user-1" },
+    where: { id: seedUserId1 },
     update: {
       name: "Demo User",
       email: "demo@langfuse.com",
       password: await hash("password", 12),
     },
     create: {
-      id: "user-1",
+      id: seedUserId1,
       name: "Demo User",
       email: "demo@langfuse.com",
       password: await hash("password", 12),
       image: "https://static.langfuse.com/langfuse-dev%2Fexample-avatar.png",
     },
   });
+  const user2 = await prisma.user.upsert({
+    where: { id: seedUserId2 },
+    update: {
+      name: "Demo User 2",
+      email: "member@langfuse.com",
+      password: await hash("password", 12),
+    },
+    create: {
+      id: seedUserId2,
+      name: "Demo User 2",
+      email: "member@langfuse.com",
+      password: await hash("password", 12),
+    },
+  });
 
-  const seedProjectId = "7a88fb47-b4e2-43b8-a06c-a5ce950dc53a";
+  await prisma.organization.upsert({
+    where: { id: seedOrgId },
+    update: {
+      name: "Seed Org",
+    },
+    create: {
+      id: seedOrgId,
+      name: "Seed Org",
+    },
+  });
 
   const project1 = await prisma.project.upsert({
     where: { id: seedProjectId },
     update: {
       name: "llm-app",
+      orgId: seedOrgId,
     },
     create: {
       id: seedProjectId,
       name: "llm-app",
-      projectMembers: {
-        create: {
-          role: "OWNER",
-          userId: user.id,
-        },
+      orgId: seedOrgId,
+    },
+  });
+
+  const orgMembership = await prisma.organizationMembership.upsert({
+    where: {
+      orgId_userId: {
+        userId: user.id,
+        orgId: seedOrgId,
       },
     },
+    create: {
+      userId: user.id,
+      orgId: seedOrgId,
+      role: "OWNER",
+    },
+    update: {},
+  });
+
+  const orgMembership2 = await prisma.organizationMembership.upsert({
+    where: {
+      orgId_userId: {
+        userId: user2.id,
+        orgId: seedOrgId,
+      },
+    },
+    create: {
+      userId: user2.id,
+      orgId: seedOrgId,
+      role: "MEMBER",
+    },
+    update: {},
+  });
+
+  const projectMembership = await prisma.projectMembership.upsert({
+    where: {
+      projectId_userId: {
+        projectId: project1.id,
+        userId: user2.id,
+      },
+    },
+    create: {
+      userId: user2.id,
+      projectId: project1.id,
+      role: "ADMIN",
+      orgMembershipId: orgMembership2.id,
+    },
+    update: {
+      orgMembershipId: orgMembership2.id,
+    },
   });
 
   await prisma.prompt.upsert({
@@ -113,18 +186,39 @@ async function main() {
 
   // Do not run the following for local docker compose setup
   if (environment === "examples" || environment === "load") {
-    const project2 = await prisma.project.upsert({
-      where: { id: "239ad00f-562f-411d-af14-831c75ddd875" },
+    const seedOrgIdOrg2 = "demo-org-id";
+    const project2Id = "239ad00f-562f-411d-af14-831c75ddd875";
+    const org2 = await prisma.organization.upsert({
+      where: { id: seedOrgIdOrg2 },
+      update: {
+        name: "Langfuse Demo",
+      },
       create: {
-        id: "239ad00f-562f-411d-af14-831c75ddd875",
+        id: seedOrgIdOrg2,
+        name: "Langfuse Demo",
+      },
+    });
+    const project2 = await prisma.project.upsert({
+      where: { id: project2Id },
+      create: {
+        id: project2Id,
         name: "demo-app",
-        projectMembers: {
-          create: {
-            role: "OWNER",
-            userId: user.id,
-          },
+        orgId: org2.id,
+      },
+      update: { orgId: seedOrgIdOrg2 },
+    });
+    await prisma.organizationMembership.upsert({
+      where: {
+        orgId_userId: {
+          userId: user.id,
+          orgId: seedOrgIdOrg2,
         },
       },
+      create: {
+        userId: user.id,
+        orgId: seedOrgIdOrg2,
+        role: "VIEWER",
+      },
       update: {},
     });
 
@@ -163,22 +257,30 @@ async function main() {
 
     const traceVolume = environment === "load" ? LOAD_TRACE_VOLUME : 100;
 
-    const { traces, observations, scores, sessions, events } = createObjects(
-      traceVolume,
-      envTags,
-      colorTags,
-      project1,
-      project2,
-      promptIds,
-      configIdsAndNames
+    const { traces, observations, scores, sessions, events, comments } =
+      createObjects(
+        traceVolume,
+        envTags,
+        colorTags,
+        project1,
+        project2,
+        promptIds,
+        configIdsAndNames,
+      );
+
+    logger.info(
+      `Seeding ${traces.length} traces, ${observations.length} observations, and ${scores.length} scores`,
     );
 
-    console.log(
-      `Seeding ${traces.length} traces, ${observations.length} observations, and ${scores.length} scores`
+    await uploadObjects(
+      traces,
+      observations,
+      scores,
+      sessions,
+      events,
+      comments,
     );
 
-    await uploadObjects(traces, observations, scores, sessions, events);
-
     // If openai key is in environment, add it to the projects LLM API keys
     const OPENAI_API_KEY = process.env.OPENAI_API_KEY;
 
@@ -193,8 +295,8 @@ async function main() {
         },
       });
     } else {
-      console.warn(
-        "No OPENAI_API_KEY found in environment. Skipping seeding LLM API key."
+      logger.warn(
+        "No OPENAI_API_KEY found in environment. Skipping seeding LLM API key.",
       );
     }
 
@@ -286,6 +388,7 @@ async function main() {
             : undefined;
         const datasetItem = await prisma.datasetItem.create({
           data: {
+            projectId: project2.id,
             datasetId: dataset.id,
             sourceTraceId: sourceObservation?.traceId,
             sourceObservationId:
@@ -312,6 +415,7 @@ async function main() {
       for (let datasetRunNumber = 0; datasetRunNumber < 5; datasetRunNumber++) {
         const datasetRun = await prisma.datasetRuns.create({
           data: {
+            projectId: project2.id,
             name: `demo-dataset-run-${datasetRunNumber}`,
             description: Math.random() > 0.5 ? "Dataset run description" : "",
             datasetId: dataset.id,
@@ -327,7 +431,7 @@ async function main() {
 
         for (const datasetItemId of datasetItemIds) {
           const relevantObservations = observations.filter(
-            (o) => o.projectId === project2.id
+            (o) => o.projectId === project2.id,
           );
           const observation =
             relevantObservations[
@@ -336,6 +440,7 @@ async function main() {
 
           await prisma.datasetRunItems.create({
             data: {
+              projectId: project2.id,
               datasetItemId,
               traceId: observation.traceId as string,
               observationId: Math.random() > 0.5 ? observation.id : undefined,
@@ -351,10 +456,14 @@ async function main() {
 main()
   .then(async () => {
     await prisma.$disconnect();
+    redis?.disconnect();
+    logger.info("Disconnected from postgres and redis");
   })
   .catch(async (e) => {
-    console.error(e);
+    logger.error(e);
     await prisma.$disconnect();
+    redis?.disconnect();
+    logger.info("Disconnected from postgres and redis");
     process.exit(1);
   });
 
@@ -363,7 +472,8 @@ async function uploadObjects(
   observations: Prisma.ObservationCreateManyInput[],
   scores: Prisma.ScoreCreateManyInput[],
   sessions: Prisma.TraceSessionCreateManyInput[],
-  events: Prisma.ObservationCreateManyInput[]
+  events: Prisma.ObservationCreateManyInput[],
+  comments: Prisma.CommentCreateManyInput[],
 ) {
   let promises: Prisma.PrismaPromise<unknown>[] = [];
 
@@ -377,14 +487,14 @@ async function uploadObjects(
         },
         create: chunk[0]!,
         update: {},
-      })
+      }),
     );
   });
 
   for (let i = 0; i < promises.length; i++) {
     if (i + 1 >= promises.length || i % Math.ceil(promises.length / 10) === 0)
-      console.log(
-        `Seeding of Sessions ${((i + 1) / promises.length) * 100}% complete`
+      logger.info(
+        `Seeding of Sessions ${((i + 1) / promises.length) * 100}% complete`,
       );
     await promises[i];
   }
@@ -395,13 +505,13 @@ async function uploadObjects(
     promises.push(
       prisma.trace.createMany({
         data: chunk,
-      })
+      }),
     );
   });
   for (let i = 0; i < promises.length; i++) {
     if (i + 1 >= promises.length || i % Math.ceil(promises.length / 10) === 0)
-      console.log(
-        `Seeding of Traces ${((i + 1) / promises.length) * 100}% complete`
+      logger.info(
+        `Seeding of Traces ${((i + 1) / promises.length) * 100}% complete`,
       );
     await promises[i];
   }
@@ -411,14 +521,14 @@ async function uploadObjects(
     promises.push(
       prisma.observation.createMany({
         data: chunk,
-      })
+      }),
     );
   });
 
   for (let i = 0; i < promises.length; i++) {
     if (i + 1 >= promises.length || i % Math.ceil(promises.length / 10) === 0)
-      console.log(
-        `Seeding of Observations ${((i + 1) / promises.length) * 100}% complete`
+      logger.info(
+        `Seeding of Observations ${((i + 1) / promises.length) * 100}% complete`,
       );
     await promises[i];
   }
@@ -428,14 +538,14 @@ async function uploadObjects(
     promises.push(
       prisma.observation.createMany({
         data: chunk,
-      })
+      }),
     );
   });
 
   for (let i = 0; i < promises.length; i++) {
     if (i + 1 >= promises.length || i % Math.ceil(promises.length / 10) === 0)
-      console.log(
-        `Seeding of Events ${((i + 1) / promises.length) * 100}% complete`
+      logger.info(
+        `Seeding of Events ${((i + 1) / promises.length) * 100}% complete`,
       );
     await promises[i];
   }
@@ -445,13 +555,29 @@ async function uploadObjects(
     promises.push(
       prisma.score.createMany({
         data: chunk,
-      })
+      }),
     );
   });
   for (let i = 0; i < promises.length; i++) {
     if (i + 1 >= promises.length || i % Math.ceil(promises.length / 10) === 0)
-      console.log(
-        `Seeding of Scores ${((i + 1) / promises.length) * 100}% complete`
+      logger.info(
+        `Seeding of Scores ${((i + 1) / promises.length) * 100}% complete`,
+      );
+    await promises[i];
+  }
+
+  promises = [];
+  chunk(comments, chunkSize).forEach((chunk) => {
+    promises.push(
+      prisma.comment.createMany({
+        data: chunk,
+      }),
+    );
+  });
+  for (let i = 0; i < promises.length; i++) {
+    if (i + 1 >= promises.length || i % Math.ceil(promises.length / 10) === 0)
+      logger.info(
+        `Seeding of Comments ${((i + 1) / promises.length) * 100}% complete`,
       );
     await promises[i];
   }
@@ -472,7 +598,7 @@ function createObjects(
       dataType: ScoreDataType;
       categories: ConfigCategory[] | null;
     }[]
-  >
+  >,
 ) {
   const traces: Prisma.TraceCreateManyInput[] = [];
   const observations: Prisma.ObservationCreateManyInput[] = [];
@@ -480,12 +606,13 @@ function createObjects(
   const sessions: Prisma.TraceSessionCreateManyInput[] = [];
   const events: Prisma.ObservationCreateManyInput[] = [];
   const configs: Prisma.ScoreConfigCreateManyInput[] = [];
+  const comments: Prisma.CommentCreateManyInput[] = [];
 
   for (let i = 0; i < traceVolume; i++) {
     // print progress to console with a progress bar that refreshes every 10 iterations
     // random date within last 90 days, with a linear bias towards more recent dates
     const traceTs = new Date(
-      Date.now() - Math.floor(Math.random() ** 1.5 * 90 * 24 * 60 * 60 * 1000)
+      Date.now() - Math.floor(Math.random() ** 1.5 * 90 * 24 * 60 * 60 * 1000),
     );
 
     const envTag = envTags[Math.floor(Math.random() * envTags.length)];
@@ -510,6 +637,7 @@ function createObjects(
     const trace = {
       id: `trace-${v4()}`,
       timestamp: traceTs,
+      createdAt: traceTs,
       projectId: projectId,
       name: ["generate-outreach", "label-inbound", "draft-response"][
         i % 3
@@ -568,6 +696,7 @@ function createObjects(
               traceId: trace.id,
               name: annotationScoreName,
               timestamp: traceTs,
+              createdAt: traceTs,
               source: ScoreSource.ANNOTATION,
               projectId,
               authorUserId: `user-${i}`,
@@ -584,6 +713,7 @@ function createObjects(
               name: "sentiment",
               value: Math.floor(Math.random() * 10) - 5,
               timestamp: traceTs,
+              createdAt: traceTs,
               source: ScoreSource.API,
               projectId,
               dataType: ScoreDataType.NUMERIC,
@@ -596,6 +726,7 @@ function createObjects(
               traceId: trace.id,
               name: "Completeness",
               timestamp: traceTs,
+              createdAt: traceTs,
               source: ScoreSource.API,
               projectId,
               dataType: ScoreDataType.CATEGORICAL,
@@ -606,6 +737,15 @@ function createObjects(
         : []),
     ];
 
+    if (Math.random() > 0.9)
+      comments.push({
+        projectId: trace.projectId,
+        objectId: trace.id,
+        objectType: "TRACE",
+        content: "Trace comment content",
+        ...(Math.random() > 0.5 ? { authorUserId: `user-${i}` } : {}),
+      });
+
     scores.push(...traceScores);
 
     const existingSpanIds: string[] = [];
@@ -613,17 +753,18 @@ function createObjects(
     for (let j = 0; j < Math.floor(Math.random() * 10) + 1; j++) {
       // add between 1 and 30 ms to trace timestamp
       const spanTsStart = new Date(
-        traceTs.getTime() + Math.floor(Math.random() * 30)
+        traceTs.getTime() + Math.floor(Math.random() * 30),
       );
       // random duration of upto 5000ms
       const spanTsEnd = new Date(
-        spanTsStart.getTime() + Math.floor(Math.random() * 5000)
+        spanTsStart.getTime() + Math.floor(Math.random() * 5000),
       );
 
       const span = {
         type: ObservationType.SPAN,
         id: `span-${v4()}`,
         startTime: spanTsStart,
+        createdAt: spanTsStart,
         endTime: spanTsEnd,
         name: `span-${i}-${j}`,
         metadata: {
@@ -651,22 +792,22 @@ function createObjects(
         const generationTsStart = new Date(
           spanTsStart.getTime() +
             Math.floor(
-              Math.random() * (spanTsEnd.getTime() - spanTsStart.getTime())
-            )
+              Math.random() * (spanTsEnd.getTime() - spanTsStart.getTime()),
+            ),
         );
         const generationTsEnd = new Date(
           generationTsStart.getTime() +
             Math.floor(
               Math.random() *
-                (spanTsEnd.getTime() - generationTsStart.getTime())
-            )
+                (spanTsEnd.getTime() - generationTsStart.getTime()),
+            ),
         );
         // somewhere in the middle
         const generationTsCompletionStart = new Date(
           generationTsStart.getTime() +
             Math.floor(
-              (generationTsEnd.getTime() - generationTsStart.getTime()) / 3
-            )
+              (generationTsEnd.getTime() - generationTsStart.getTime()) / 3,
+            ),
         );
 
         const promptTokens = Math.floor(Math.random() * 1000) + 300;
@@ -687,49 +828,25 @@ function createObjects(
         const promptId =
           promptIds.get(projectId)![
             Math.floor(
-              Math.random() * Math.floor(promptIds.get(projectId)!.length / 2)
+              Math.random() * Math.floor(promptIds.get(projectId)!.length / 2),
             )
           ];
 
+        const { input, output } = getGenerationInputOutput();
+
         const generation = {
           type: ObservationType.GENERATION,
           id: `generation-${v4()}`,
           startTime: generationTsStart,
+          createdAt: generationTsStart,
           endTime: generationTsEnd,
           completionStartTime:
             Math.random() > 0.5 ? generationTsCompletionStart : undefined,
           name: `generation-${i}-${j}-${k}`,
           projectId: trace.projectId,
           promptId: promptId,
-          input:
-            Math.random() > 0.5
-              ? [
-                  {
-                    role: "system",
-                    content: "Be a helpful assistant",
-                  },
-                  {
-                    role: "user",
-                    content: "How can i create a *React* component?",
-                  },
-                ]
-              : {
-                  input: "How can i create a React component?",
-                  retrievedDocuments: [
-                    {
-                      title: "How to create a React component",
-                      url: "https://www.google.com",
-                      description: "A guide to creating React components",
-                    },
-                    {
-                      title: "React component creation",
-                      url: "https://www.google.com",
-                      description: "A guide to creating React components",
-                    },
-                  ],
-                },
-          output:
-            "Creating a React component can be done in two ways: as a functional component or as a class component. Let's start with a basic example of both.\n\n1.  **Functional Component**:\n\nA functional component is just a plain JavaScript function that accepts props as an argument, and returns a React element. Here's how you can create one:\n\n```javascript\nimport React from 'react';\nfunction Greeting(props) {\n  return <h1>Hello, {props.name}</h1>;\n}\nexport default Greeting;\n```\n\nTo use this component in another file, you can do:\n\n```javascript\nimport Greeting from './Greeting';\nfunction App() {\n  return (\n    <div>\n      <Greeting name=\"John\" />\n    </div>\n  );\n}\nexport default App;\n```\n\n2.  **Class Component**:\n\nYou can also define components as classes in React. These have some additional features compared to functional components:\n\n```javascript\nimport React, { Component } from 'react';\nclass Greeting extends Component {\n  render() {\n    return <h1>Hello, {this.props.name}</h1>;\n  }\n}\nexport default Greeting;\n```\n\nAnd here's how to use this component:\n\n```javascript\nimport Greeting from './Greeting';\nclass App extends Component {\n  render() {\n    return (\n      <div>\n        <Greeting name=\"John\" />\n      </div>\n    );\n  }\n}\nexport default App;\n```\n\nWith the advent of hooks in React, functional components can do everything that class components can do and hence, the community has been favoring functional components over class components.\n\nRemember to import React at the top of your file whenever you're creating a component, because JSX transpiles to `React.createElement` calls under the hood.",
+          input,
+          output,
           model: model,
           internalModel: model,
           modelParameters: {
@@ -765,6 +882,8 @@ function createObjects(
             traceId: trace.id,
             source: ScoreSource.API,
             projectId: trace.projectId,
+            timestamp: generationTsEnd,
+            createdAt: traceTs,
           });
         if (Math.random() > 0.6)
           scores.push({
@@ -774,6 +893,16 @@ function createObjects(
             traceId: trace.id,
             source: ScoreSource.API,
             projectId: trace.projectId,
+            timestamp: generationTsEnd,
+            createdAt: traceTs,
+          });
+
+        if (Math.random() > 0.8)
+          comments.push({
+            projectId: trace.projectId,
+            objectId: generation.id,
+            objectType: "OBSERVATION",
+            content: "Observation comment content",
           });
 
         for (let l = 0; l < Math.floor(Math.random() * 2); l++) {
@@ -781,14 +910,15 @@ function createObjects(
           const eventTs = new Date(
             spanTsStart.getTime() +
               Math.floor(
-                Math.random() * (spanTsEnd.getTime() - spanTsStart.getTime())
-              )
+                Math.random() * (spanTsEnd.getTime() - spanTsStart.getTime()),
+              ),
           );
 
           events.push({
             type: ObservationType.EVENT,
             id: `event-${v4()}`,
             startTime: eventTs,
+            createdAt: eventTs,
             name: `event-${i}-${j}-${k}-${l}`,
             metadata: {
               user: `user-${i}@langfuse.com`,
@@ -803,7 +933,7 @@ function createObjects(
   }
   // find unique sessions by id and projectid
   const uniqueSessions: Prisma.TraceSessionCreateManyInput[] = Array.from(
-    new Set(sessions.map((session) => JSON.stringify(session)))
+    new Set(sessions.map((session) => JSON.stringify(session))),
   ).map((session) => JSON.parse(session) as Prisma.TraceSessionCreateManyInput);
 
   return {
@@ -813,6 +943,7 @@ function createObjects(
     configs,
     sessions: uniqueSessions,
     events,
+    comments,
   };
 }
 
@@ -823,7 +954,7 @@ async function generatePromptsForProject(projects: Project[]) {
     projects.map(async (project) => {
       const promptIdsForProject = await generatePrompts(project);
       promptIds.set(project.id, promptIdsForProject);
-    })
+    }),
   );
   return promptIds;
 }
@@ -1009,7 +1140,7 @@ async function generateConfigsForProject(projects: Project[]) {
     projects.map(async (project) => {
       const configNameAndId = await generateConfigs(project);
       projectIdsToConfigs.set(project.id, configNameAndId);
-    })
+    }),
   );
   return projectIdsToConfigs;
 }
@@ -1087,3 +1218,63 @@ async function generateConfigs(project: Project) {
 
   return configNameAndId;
 }
+function getGenerationInputOutput(): {
+  input: Prisma.InputJsonValue;
+  output: Prisma.InputJsonValue;
+} {
+  if (Math.random() > 0.9) {
+    const input = [
+      {
+        role: "user",
+        content: [
+          { text: "What’s depicted in this image?", type: "text" },
+          {
+            type: "image_url",
+            image_url: {
+              url: "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg",
+            },
+          },
+          { text: "Describe the scene in detail.", type: "text" },
+        ],
+      },
+    ];
+
+    const output =
+      "The image depicts a serene landscape featuring a wooden pathway or boardwalk that winds through a lush green field. The field is filled with tall grass and surrounded by trees and shrubs. Above, the sky is bright with scattered clouds, suggesting a clear and pleasant day. The scene conveys a sense of tranquility and natural beauty.";
+
+    return { input, output };
+  }
+
+  const input =
+    Math.random() > 0.5
+      ? [
+          {
+            role: "system",
+            content: "Be a helpful assistant",
+          },
+          {
+            role: "user",
+            content: "How can i create a *React* component?",
+          },
+        ]
+      : {
+          input: "How can i create a React component?",
+          retrievedDocuments: [
+            {
+              title: "How to create a React component",
+              url: "https://www.google.com",
+              description: "A guide to creating React components",
+            },
+            {
+              title: "React component creation",
+              url: "https://www.google.com",
+              description: "A guide to creating React components",
+            },
+          ],
+        };
+
+  const output =
+    "Creating a React component can be done in two ways: as a functional component or as a class component. Let's start with a basic example of both.\n\n**Image**\n\n![Languse Example Image](https://static.langfuse.com/langfuse-dev/langfuse-example-image.jpeg)\n\n1.  **Functional Component**:\n\nA functional component is just a plain JavaScript function that accepts props as an argument, and returns a React element. Here's how you can create one:\n\n```javascript\nimport React from 'react';\nfunction Greeting(props) {\n  return <h1>Hello, {props.name}</h1>;\n}\nexport default Greeting;\n```\n\nTo use this component in another file, you can do:\n\n```javascript\nimport Greeting from './Greeting';\nfunction App() {\n  return (\n    <div>\n      <Greeting name=\"John\" />\n    </div>\n  );\n}\nexport default App;\n```\n\n2.  **Class Component**:\n\nYou can also define components as classes in React. These have some additional features compared to functional components:\n\n```javascript\nimport React, { Component } from 'react';\nclass Greeting extends Component {\n  render() {\n    return <h1>Hello, {this.props.name}</h1>;\n  }\n}\nexport default Greeting;\n```\n\nAnd here's how to use this component:\n\n```javascript\nimport Greeting from './Greeting';\nclass App extends Component {\n  render() {\n    return (\n      <div>\n        <Greeting name=\"John\" />\n      </div>\n    );\n  }\n}\nexport default App;\n```\n\nWith the advent of hooks in React, functional components can do everything that class components can do and hence, the community has been favoring functional components over class components.\n\nRemember to import React at the top of your file whenever you're creating a component, because JSX transpiles to `React.createElement` calls under the hood.";
+
+  return { input, output };
+}

packages/shared/src/constants.ts

@@ -6,4 +6,5 @@ export enum ModelUsageUnit {
   Seconds = "SECONDS",
   Milliseconds = "MILLISECONDS",
   Images = "IMAGES",
+  Requests = "REQUESTS",
 }

packages/shared/src/db.ts

@@ -1,7 +1,7 @@
 // This file exports the prisma db connection, the Prisma Object, and the Typescript types.
 // This is not imported in the index.ts file of this package, as we must not import this into FE code.
 
-import { PrismaClient } from "@prisma/client";
+import { Prisma, PrismaClient } from "@prisma/client";
 import { env } from "process";
 import kyselyExtension from "prisma-extension-kysely";
 import {
@@ -11,17 +11,38 @@ import {
   PostgresQueryCompiler,
 } from "kysely";
 import { DB } from ".";
+import { logger } from "./server";
 
 // Instantiated according to the Prisma documentation
 // https://www.prisma.io/docs/orm/more/help-and-troubleshooting/help-articles/nextjs-prisma-client-dev-practices
 
 const prismaClientSingleton = () => {
-  return new PrismaClient({
-    log:
-      env.NODE_ENV === "development"
-        ? ["query", "error", "warn"]
-        : ["error", "warn"],
+  const client = new PrismaClient<
+    Prisma.PrismaClientOptions,
+    "warn" | "error" | "query"
+  >({
+    log: [
+      { emit: "event", level: "query" },
+      { emit: "event", level: "error" },
+      { emit: "event", level: "warn" },
+    ],
   });
+
+  if (env.NODE_ENV === "development") {
+    client.$on("query", (event) => {
+      logger.info(`prisma:query ${event.query}, ${event.duration}ms`);
+    });
+  }
+
+  client.$on("warn", (event) => {
+    logger.warn(`prisma:warn ${event.message}`);
+  });
+
+  client.$on("error", (event) => {
+    logger.error(`prisma:error ${event.message}`);
+  });
+
+  return client;
 };
 
 const kyselySingleton = (prismaClient: PrismaClient) => {
@@ -38,7 +59,7 @@ const kyselySingleton = (prismaClient: PrismaClient) => {
             createQueryCompiler: () => new PostgresQueryCompiler(),
           },
         }),
-    })
+    }),
   );
 };
 declare global {

packages/shared/src/encryption/index.ts

@@ -1,7 +1,7 @@
 import crypto from "crypto";
 import { env } from "../env";
 
-const ENCRYPTION_KEY: string = env.ENCRYPTION_KEY || ""; // Must be 256 bits (32 bytes, 64 hex characters)
+const ENCRYPTION_KEY: string | undefined = env.ENCRYPTION_KEY; // Must be 256 bits (32 bytes, 64 hex characters)
 const IV_LENGTH: number = 16; // For AES, this is always 16
 
 // Alternatively: openssl rand -hex 32
@@ -13,17 +13,16 @@ export function keyGen() {
  * Encrypts the given plain text using AES-256-GCM algorithm.
  *
  * @param {string} plainText - The text to encrypt.
- * @param {string} [keyHex=ENCRYPTION_KEY] - The encryption key in hexadecimal format.
  * @returns {string} The encrypted data in hex format, including IV and authentication tag.
  */
-export function encrypt(
-  plainText: string,
-  keyHex: string = ENCRYPTION_KEY
-): string {
+export function encrypt(plainText: string): string {
+  if (!ENCRYPTION_KEY) {
+    throw new Error("Missing environment variable: `ENCRYPTION_KEY`");
+  }
   const iv = crypto.randomBytes(IV_LENGTH); // Directly use Buffer returned by randomBytes
   const cipher = crypto.createCipheriv(
     "aes-256-gcm",
-    Buffer.from(keyHex, "hex"),
+    Buffer.from(ENCRYPTION_KEY, "hex"),
     iv
   );
   let encrypted = cipher.update(plainText, "utf8", "hex");
@@ -34,7 +33,10 @@ export function encrypt(
   return iv.toString("hex") + ":" + encrypted + ":" + authTag.toString("hex");
 }
 
-export function decrypt(text: string, keyHex: string = ENCRYPTION_KEY): string {
+export function decrypt(text: string): string {
+  if (!ENCRYPTION_KEY) {
+    throw new Error("Missing environment variable: `ENCRYPTION_KEY`");
+  }
   const [ivHex, encryptedHex, authTagHex] = text.split(":");
   if (!ivHex || !encryptedHex || !authTagHex) {
     throw new Error("Invalid or corrupted cipher format");
@@ -46,7 +48,7 @@ export function decrypt(text: string, keyHex: string = ENCRYPTION_KEY): string {
 
   const decipher = crypto.createDecipheriv(
     "aes-256-gcm",
-    Buffer.from(keyHex, "hex"),
+    Buffer.from(ENCRYPTION_KEY, "hex"),
     iv
   );
   decipher.setAuthTag(authTag);

packages/shared/src/env.ts

@@ -1,13 +1,48 @@
 import { z } from "zod";
 
 const EnvSchema = z.object({
+  NODE_ENV: z
+    .enum(["development", "test", "production"])
+    .default("development"),
+  REDIS_HOST: z.string().nullish(),
+  REDIS_PORT: z.coerce
+    .number({
+      description:
+        ".env files convert numbers to strings, therefoore we have to enforce them to be numbers",
+    })
+    .positive()
+    .max(65536, `options.port should be >= 0 and < 65536`)
+    .default(6379)
+    .nullable(),
+  REDIS_AUTH: z.string().nullish(),
+  REDIS_CONNECTION_STRING: z.string().nullish(),
+  REDIS_ENABLE_AUTO_PIPELINING: z.enum(["true", "false"]).default("true"),
   ENCRYPTION_KEY: z
     .string()
     .length(
       64,
-      "ENCRYPTION_KEY must be 256 bits, 64 string characters in hex format, generate via: openssl rand -hex 32"
+      "ENCRYPTION_KEY must be 256 bits, 64 string characters in hex format, generate via: openssl rand -hex 32",
     )
     .optional(),
+  LANGFUSE_CACHE_PROMPT_ENABLED: z.enum(["true", "false"]).default("false"),
+  LANGFUSE_CACHE_PROMPT_TTL_SECONDS: z.coerce.number().default(60 * 60),
+  CLICKHOUSE_URL: z.string().url().optional(),
+  CLICKHOUSE_USER: z.string().optional(),
+  CLICKHOUSE_PASSWORD: z.string().optional(),
+  LANGFUSE_INGESTION_FLUSH_DELAY_MS: z.coerce
+    .number()
+    .nonnegative()
+    .default(10000),
+  LANGFUSE_INGESTION_FLUSH_ATTEMPTS: z.coerce.number().positive().default(3),
+  LANGFUSE_INGESTION_BUFFER_TTL_SECONDS: z.coerce
+    .number()
+    .positive()
+    .default(60 * 10),
+  SALT: z.string().optional(), // used by components imported by web package
+  LANGFUSE_LOG_LEVEL: z
+    .enum(["trace", "debug", "info", "warn", "error", "fatal"])
+    .optional(),
+  LANGFUSE_LOG_FORMAT: z.enum(["text", "json"]).default("text"),
 });
 
 export const env = EnvSchema.parse(process.env);

packages/shared/src/features/annotation/types.ts

@@ -14,7 +14,7 @@ const NumericData = z.object({
 });
 
 const CategoricalData = z.object({
-  value: z.number().optional().nullish(),
+  value: z.number().nullish(),
   stringValue: z.string(),
   dataType: z.literal("CATEGORICAL"),
 });
@@ -31,7 +31,7 @@ const CreateAnnotationScoreBase = z.object({
   traceId: z.string(),
   configId: z.string().optional(),
   observationId: z.string().optional(),
-  comment: z.string().optional().nullish(),
+  comment: z.string().nullish(),
 });
 
 const UpdateAnnotationScoreBase = CreateAnnotationScoreBase.extend({

packages/shared/src/features/comments/types.ts

@@ -0,0 +1,22 @@
+import { z } from "zod";
+
+const COMMENT_OBJECT_TYPES = [
+  "TRACE",
+  "OBSERVATION",
+  "SESSION",
+  "PROMPT",
+] as const;
+
+export const CreateCommentData = z.object({
+  projectId: z.string(),
+  content: z.string().trim().min(1).max(500),
+  objectId: z.string(),
+  objectType: z.enum(COMMENT_OBJECT_TYPES),
+});
+
+export const DeleteCommentData = z.object({
+  projectId: z.string(),
+  commentId: z.string(),
+  objectId: z.string(),
+  objectType: z.enum(COMMENT_OBJECT_TYPES),
+});

packages/shared/src/features/entitlements/plans.ts

@@ -0,0 +1,17 @@
+export const planLabels = {
+  oss: "OSS",
+  "cloud:hobby": "Hobby",
+  "cloud:pro": "Pro",
+  "cloud:team": "Team",
+  "self-hosted:enterprise": "Enterprise",
+} as const;
+
+export type Plan = keyof typeof planLabels;
+
+export const plans = Object.keys(planLabels) as Plan[];
+
+// This function is kept here to ensure consistency when updating plan names in the future.
+export const isCloudPlan = (plan: Plan) => plan.startsWith("cloud");
+
+export const isPlan = (value: string): value is Plan =>
+  plans.includes(value as Plan);

packages/shared/src/features/evals/types.ts

@@ -80,6 +80,8 @@ export const availableEvalVariables = [
 export const evalModelsByAdapter: Record<LLMAdapter, string[]> = {
   [LLMAdapter.OpenAI]: [
     "gpt-4o",
+    "gpt-4o-mini",
+    "gpt-4o-2024-08-06", // cheaper than gpt-4o
     "gpt-4-turbo-preview",
     "gpt-3.5-turbo",
   ] as const,

packages/shared/src/features/scores/scoreConfigTypes.ts

@@ -1,12 +1,13 @@
-import { isPresent } from "@/src/utils/typeChecks";
+import { z } from "zod";
+
+import { ScoreConfig as ScoreConfigDbType } from "@prisma/client";
+
+import { isPresent } from "../../utils/typeChecks";
 import {
   jsonSchema,
   paginationMetaResponseZod,
   paginationZod,
-  type ScoreConfig as ScoreConfigDbType,
-} from "@langfuse/shared";
-import { z } from "zod";
-import * as Sentry from "@sentry/node";
+} from "../../utils/zod";
 
 /**
  * Types to use across codebase
@@ -16,7 +17,7 @@ export type ValidatedScoreConfig = z.infer<typeof ValidatedScoreConfigSchema>;
 
 const validateCategories = (
   categories: ConfigCategory[],
-  ctx: z.RefinementCtx,
+  ctx: z.RefinementCtx
 ) => {
   const uniqueNames = new Set<string>();
   const uniqueValues = new Set<number>();
@@ -53,8 +54,8 @@ export const Category = z.object({
 const Categories = z.array(Category);
 
 const NumericScoreConfig = z.object({
-  maxValue: z.number().optional().nullish(),
-  minValue: z.number().optional().nullish(),
+  maxValue: z.number().nullish(),
+  minValue: z.number().nullish(),
   dataType: z.literal("NUMERIC"),
   categories: z.undefined().nullish(),
 });
@@ -93,7 +94,7 @@ const BooleanScoreConfig = z.object({
       return categories.every(
         (category, index) =>
           category.label === expectedCategories[index].label &&
-          category.value === expectedCategories[index].value,
+          category.value === expectedCategories[index].value
       );
     }),
 });
@@ -102,7 +103,7 @@ const ScoreConfigBase = z.object({
   id: z.string(),
   name: z.string().min(1).max(35),
   isArchived: z.boolean(),
-  description: z.string().optional().nullish(),
+  description: z.string().nullish(),
   createdAt: z.coerce.date(),
   updatedAt: z.coerce.date(),
   projectId: z.string(),
@@ -122,7 +123,7 @@ const ValidatedScoreConfigSchema = z
         minValue: z.undefined().nullish(),
         dataType: z.literal("CATEGORICAL"),
         categories: Categories.superRefine(validateCategories),
-      }),
+      })
     ),
     ScoreConfigBase.merge(BooleanScoreConfig),
   ])
@@ -143,19 +144,20 @@ const ValidatedScoreConfigSchema = z
 
 /**
  * Use this function when pulling a list of score configs from the database before using in the application to ensure type safety.
- * All score configs are expected to pass the validation. If a score fails validation, it will be logged to Sentry.
+ * All score configs are expected to pass the validation. If a score fails validation, it will be logged to Otel.
  * @param scoreConfigs
  * @returns list of validated score configs
  */
 export const filterAndValidateDbScoreConfigList = (
   scoreConfigs: ScoreConfigDbType[],
+  onParseError?: (error: z.ZodError) => void
 ): ValidatedScoreConfig[] =>
   scoreConfigs.reduce((acc, ts) => {
     const result = ValidatedScoreConfigSchema.safeParse(ts);
     if (result.success) {
       acc.push(result.data);
     } else {
-      Sentry.captureException(result.error);
+      onParseError?.(result.error);
     }
     return acc;
   }, [] as ValidatedScoreConfig[]);
@@ -168,7 +170,7 @@ export const filterAndValidateDbScoreConfigList = (
  * @throws error if score fails validation
  */
 export const validateDbScoreConfig = (
-  scoreConfig: ScoreConfigDbType,
+  scoreConfig: ScoreConfigDbType
 ): ValidatedScoreConfig => ValidatedScoreConfigSchema.parse(scoreConfig);
 
 /**
@@ -203,7 +205,7 @@ export const PostScoreConfigBody = z
       z.object({
         dataType: z.literal("BOOLEAN"),
         categories: z.undefined().nullish(),
-      }),
+      })
     ),
   ])
   .superRefine((data, ctx) => {

packages/shared/src/features/scores/scoreTypes.ts

@@ -1,19 +1,19 @@
-import * as Sentry from "@sentry/node";
-import {
-  paginationZod,
-  paginationMetaResponseZod,
-  NonEmptyString,
-  type Score,
-  stringDateTime,
-} from "@langfuse/shared";
 import { z } from "zod";
-import { isPresent } from "@/src/utils/typeChecks";
-import { Category as ConfigCategory } from "./score-configs";
+
+import { Score } from "@prisma/client";
+
+import { isPresent, stringDateTime } from "../../utils/typeChecks";
+import {
+  NonEmptyString,
+  paginationMetaResponseZod,
+  paginationZod,
+} from "../../utils/zod";
+import { Category as ConfigCategory } from "./scoreConfigTypes";
 
 /**
  * Types to use across codebase
  */
-export type APIScore = z.infer<typeof APIScore>;
+export type APIScore = z.infer<typeof APIScoreSchema>;
 
 /**
  * Helpers
@@ -31,7 +31,7 @@ const NumericData = z.object({
 });
 
 const CategoricalData = z.object({
-  value: z.number().optional().nullish(),
+  value: z.number().nullish(),
   stringValue: z.string(),
   dataType: z.literal("CATEGORICAL"),
 });
@@ -69,7 +69,7 @@ const BaseScoreBody = z.object({
  * Objects
  */
 
-export const APIScore = z.discriminatedUnion("dataType", [
+export const APIScoreSchema = z.discriminatedUnion("dataType", [
   ScoreBase.merge(NumericData),
   ScoreBase.merge(CategoricalData),
   ScoreBase.merge(BooleanData),
@@ -155,18 +155,21 @@ export const ScorePropsAgainstConfig = z.union([
 
 /**
  * Use this function when pulling a list of scores from the database before using in the application to ensure type safety.
- * All scores are expected to pass the validation. If a score fails validation, it will be logged to Sentry.
+ * All scores are expected to pass the validation. If a score fails validation, it will be logged to Otel.
  * @param scores
  * @returns list of validated scores
  */
-export const filterAndValidateDbScoreList = (scores: Score[]): APIScore[] =>
+export const filterAndValidateDbScoreList = (
+  scores: Score[],
+  onParseError?: (error: z.ZodError) => void,
+): APIScore[] =>
   scores.reduce((acc, ts) => {
-    const result = APIScore.safeParse(ts);
+    const result = APIScoreSchema.safeParse(ts);
     if (result.success) {
       acc.push(result.data);
     } else {
       console.error("Score parsing error: ", result.error);
-      Sentry.captureException(result.error);
+      onParseError?.(result.error);
     }
     return acc;
   }, [] as APIScore[]);
@@ -179,7 +182,7 @@ export const filterAndValidateDbScoreList = (scores: Score[]): APIScore[] =>
  * @throws error if score fails validation
  */
 export const validateDbScore = (score: Score): APIScore =>
-  APIScore.parse(score);
+  APIScoreSchema.parse(score);
 
 /**
  * Endpoints
@@ -189,63 +192,39 @@ export const validateDbScore = (score: Score): APIScore =>
 /**
  * PostScoresBody is copied for the ingestion API as `ScoreBody`. Please copy any changes here in `packages/shared/src/features/ingestion/types.ts`
  */
-export const PostScoresBody = z
-  .discriminatedUnion("dataType", [
-    BaseScoreBody.merge(
-      z.object({
-        value: z.number(),
-        dataType: z.literal("NUMERIC"),
-        configId: z.string().nullish(),
+export const PostScoresBody = z.discriminatedUnion("dataType", [
+  BaseScoreBody.merge(
+    z.object({
+      value: z.number(),
+      dataType: z.literal("NUMERIC"),
+      configId: z.string().nullish(),
+    }),
+  ),
+  BaseScoreBody.merge(
+    z.object({
+      value: z.string(),
+      dataType: z.literal("CATEGORICAL"),
+      configId: z.string().nullish(),
+    }),
+  ),
+  BaseScoreBody.merge(
+    z.object({
+      value: z.number().refine((value) => value === 0 || value === 1, {
+        message:
+          "Value must be a number equal to either 0 or 1 for data type BOOLEAN",
       }),
-    ),
-    BaseScoreBody.merge(
-      z.object({
-        value: z.string(),
-        dataType: z.literal("CATEGORICAL"),
-        configId: z.string().nullish(),
-      }),
-    ),
-    BaseScoreBody.merge(
-      z.object({
-        value: z.number().refine((val) => val === 0 || val === 1, {
-          message: "Value must be either 0 or 1",
-        }),
-        dataType: z.literal("BOOLEAN"),
-        configId: z.string().nullish(),
-      }),
-    ),
-    BaseScoreBody.merge(
-      z.object({
-        value: z.union([z.string(), z.number()]),
-        dataType: z.undefined(),
-        configId: z.string().nullish(),
-      }),
-    ),
-  ])
-  .superRefine((data, ctx) => {
-    if (data.dataType) {
-      if (typeof data.value === "number") {
-        if (data.dataType === "CATEGORICAL") {
-          ctx.addIssue({
-            code: z.ZodIssueCode.custom,
-            message: `Value must be a string for data type ${data.dataType}`,
-          });
-        }
-      } else if (typeof data.value === "string") {
-        if (data.dataType === "NUMERIC") {
-          ctx.addIssue({
-            code: z.ZodIssueCode.custom,
-            message: `Value must be a number for data type ${data.dataType}`,
-          });
-        } else if (data.dataType === "BOOLEAN") {
-          ctx.addIssue({
-            code: z.ZodIssueCode.custom,
-            message: `Value must number equal to either 0 or 1 for data type ${data.dataType}`,
-          });
-        }
-      }
-    }
-  });
+      dataType: z.literal("BOOLEAN"),
+      configId: z.string().nullish(),
+    }),
+  ),
+  BaseScoreBody.merge(
+    z.object({
+      value: z.union([z.string(), z.number()]),
+      dataType: z.undefined(),
+      configId: z.string().nullish(),
+    }),
+  ),
+]);
 
 export const PostScoresResponse = z.object({ id: z.string() });
 
@@ -272,7 +251,7 @@ export const GetScoresQuery = z.object({
 
 // LegacyGetScoreResponseDataV1 is only used for response of GET /scores list endpoint
 const LegacyGetScoreResponseDataV1 = z.intersection(
-  APIScore,
+  APIScoreSchema,
   z.object({
     trace: z.object({
       userId: z.string().nullish(),
@@ -286,6 +265,7 @@ export const GetScoresResponse = z.object({
 
 export const legacyFilterAndValidateV1GetScoreList = (
   scores: unknown[],
+  onParseError?: (error: z.ZodError) => void,
 ): z.infer<typeof LegacyGetScoreResponseDataV1>[] =>
   scores.reduce(
     (acc: z.infer<typeof LegacyGetScoreResponseDataV1>[], ts) => {
@@ -294,7 +274,7 @@ export const legacyFilterAndValidateV1GetScoreList = (
         acc.push(result.data);
       } else {
         console.error("Score parsing error: ", result.error);
-        Sentry.captureException(result.error);
+        onParseError?.(result.error);
       }
       return acc;
     },
@@ -306,7 +286,7 @@ export const GetScoreQuery = z.object({
   scoreId: z.string(),
 });
 
-export const GetScoreResponse = APIScore;
+export const GetScoreResponse = APIScoreSchema;
 
 // DELETE /scores/{scoreId}
 export const DeleteScoreQuery = z.object({

packages/shared/src/index.ts

@@ -1,18 +1,19 @@
 export * from "./constants";
-export * from "./queries";
-export * from "./queues";
 export * from "./interfaces/filters";
 export * from "./interfaces/orderBy";
+export * from "./interfaces/cloudConfigSchema";
+export * from "./interfaces/parseDbOrg";
 export * from "./tableDefinitions";
 export * from "./types";
-export * from "./filterToPrisma";
-export * from "./orderByToPrisma";
 export * from "./tracesTable";
-export * from "./server/auth";
+export * from "./server/auth/apiKeys";
 export * from "./observationsTable";
-export * from "./features/ingestion/types";
 export * from "./utils/zod";
 export * from "./utils/json";
+export * from "./utils/objects";
+export * from "./utils/typeChecks";
+export * from "./features/entitlements/plans";
+export * from "./interfaces/rate-limits";
 export { env } from "./env";
 
 // llm api
@@ -26,6 +27,13 @@ export * from "./features/batchExport/types";
 // annotation
 export * from "./features/annotation/types";
 
+// scores
+export * from "./features/scores/scoreConfigTypes";
+export * from "./features/scores/scoreTypes";
+
+// comments
+export * from "./features/comments/types";
+
 // export db types only
 export * from "@prisma/client";
 export { type DB } from "../prisma/generated/types";

packages/shared/src/interfaces/cloudConfigSchema.ts

@@ -0,0 +1,22 @@
+import { z } from "zod";
+import { CloudConfigRateLimit } from "./rate-limits";
+
+export const CloudConfigSchema = z.object({
+  plan: z.enum(["Hobby", "Pro", "Team", "Enterprise"]).optional(),
+  monthlyObservationLimit: z.number().int().positive().optional(),
+  // used for table and dashboard queries
+  defaultLookBackDays: z.number().int().positive().optional(),
+  // need to update stripe webhook if you change this, it fetches from db via these fields
+  stripe: z
+    .object({
+      customerId: z.string().optional(),
+      activeSubscriptionId: z.string().optional(),
+      activeProductId: z.string().optional(),
+    })
+    .optional(),
+
+  // custom rate limits for an organization
+  rateLimitOverrides: CloudConfigRateLimit.optional(),
+});
+
+export type CloudConfigSchema = z.infer<typeof CloudConfigSchema>;

packages/shared/src/interfaces/parseDbOrg.ts

@@ -0,0 +1,19 @@
+import { type Organization } from "@prisma/client";
+import { CloudConfigSchema } from "./cloudConfigSchema";
+
+type parsedOrg = Omit<Organization, "cloudConfig"> & {
+  cloudConfig: CloudConfigSchema | null;
+};
+
+export function parseDbOrg(dbOrg: Organization): parsedOrg {
+  const { cloudConfig, ...org } = dbOrg;
+
+  const parsedCloudConfig = CloudConfigSchema.safeParse(cloudConfig);
+
+  const parsedOrg = {
+    ...org,
+    cloudConfig: parsedCloudConfig.success ? parsedCloudConfig.data : null,
+  };
+
+  return parsedOrg;
+}

packages/shared/src/interfaces/rate-limits.ts

@@ -0,0 +1,30 @@
+import z from "zod";
+import { ApiAccessScope } from "../server";
+
+export const RateLimitResource = z.enum([
+  "ingestion",
+  "public-api",
+  "public-api-metrics",
+  "prompts",
+]);
+
+// result of the rate limit check.
+export type RateLimitResult = {
+  resource: z.infer<typeof RateLimitResource>;
+  points: number;
+  scope: ApiAccessScope;
+
+  // from rate-limiter-flexible
+  remainingPoints: number;
+  msBeforeNext: number;
+  consumedPoints: number;
+  isFirstInDuration: boolean;
+};
+
+export const RateLimitConfig = z.object({
+  points: z.number().nullish(),
+  durationInSec: z.number().nullish(),
+  resource: RateLimitResource,
+});
+
+export const CloudConfigRateLimit = z.array(RateLimitConfig);

packages/shared/src/observationsTable.ts

@@ -18,6 +18,7 @@ export const observationsTableCols: ColumnDefinition[] = [
     type: "stringOptions",
     internal: 'o."name"',
     options: [], // to be added at runtime
+    nullable: true,
   },
   { name: "Trace ID", id: "traceId", type: "string", internal: 't."id"' },
   {
@@ -26,8 +27,15 @@ export const observationsTableCols: ColumnDefinition[] = [
     type: "stringOptions",
     internal: 't."name"',
     options: [], // to be added at runtime
+    nullable: true,
+  },
+  {
+    name: "User ID",
+    id: "userId",
+    type: "string",
+    internal: 't."user_id"',
+    nullable: true,
   },
-  { name: "User ID", id: "userId", type: "string", internal: 't."user_id"' },
   {
     name: "Start Time",
     id: "startTime",
@@ -45,6 +53,7 @@ export const observationsTableCols: ColumnDefinition[] = [
     id: "timeToFirstToken",
     type: "number",
     internal: 'o."completion_start_time" - o."start_time"',
+    nullable: true,
   },
   {
     name: "Latency (s)",
@@ -57,24 +66,28 @@ export const observationsTableCols: ColumnDefinition[] = [
     id: "timePerOutputToken",
     type: "number",
     internal: '"latency" / o."completion_tokens"',
+    nullable: true,
   },
   {
     name: "Input Cost ($)",
     id: "inputCost",
     type: "number",
     internal: 'o."calculated_input_cost"',
+    nullable: true,
   },
   {
     name: "Output Cost ($)",
     id: "outputCost",
     type: "number",
     internal: 'o."calculated_output_cost"',
+    nullable: true,
   },
   {
     name: "Total Cost ($)",
     id: "totalCost",
     type: "number",
     internal: 'o."calculated_total_cost"',
+    nullable: true,
   },
   {
     name: "Level",
@@ -88,6 +101,7 @@ export const observationsTableCols: ColumnDefinition[] = [
     id: "statusMessage",
     type: "string",
     internal: 'o."status_message"',
+    nullable: true,
   },
   {
     name: "Model",
@@ -95,24 +109,28 @@ export const observationsTableCols: ColumnDefinition[] = [
     type: "stringOptions",
     internal: 'o."model"',
     options: [], // to be added at runtime
+    nullable: true,
   },
   {
     name: "Input Tokens",
     id: "inputTokens",
     type: "number",
     internal: 'o."prompt_tokens"',
+    nullable: true,
   },
   {
     name: "Output Tokens",
     id: "outputTokens",
     type: "number",
     internal: 'o."completion_tokens"',
+    nullable: true,
   },
   {
     name: "Total Tokens",
     id: "totalTokens",
     type: "number",
     internal: 'o."total_tokens"',
+    nullable: true,
   },
   {
     name: "Usage",
@@ -137,6 +155,7 @@ export const observationsTableCols: ColumnDefinition[] = [
     id: "version",
     type: "string",
     internal: 'o."version"',
+    nullable: true,
   },
   {
     name: "Prompt Name",
@@ -144,12 +163,21 @@ export const observationsTableCols: ColumnDefinition[] = [
     type: "stringOptions",
     internal: "p.name",
     options: [], // to be added at runtime
+    nullable: true,
   },
   {
     name: "Prompt Version",
     id: "promptVersion",
     type: "number",
     internal: "p.version",
+    nullable: true,
+  },
+  {
+    name: "Trace Tags",
+    id: "tags",
+    type: "arrayOptions",
+    internal: "t.tags",
+    options: [], // to be added at runtime
   },
 ];
 
@@ -161,6 +189,7 @@ export type ObservationOptions = {
   traceName: Array<OptionsDefinition>;
   scores_avg: Array<string>;
   promptName: Array<OptionsDefinition>;
+  tags: Array<OptionsDefinition>;
 };
 
 export function observationsTableColsWithOptions(
@@ -182,6 +211,9 @@ export function observationsTableColsWithOptions(
     if (col.id === "promptName") {
       return { ...col, options: options?.promptName ?? [] };
     }
+    if (col.id === "tags") {
+      return { ...col, options: options?.tags ?? [] };
+    }
     return col;
   });
 }

packages/shared/src/queries/index.ts

@@ -1 +0,0 @@
-export { getSessionTableSQL } from "./getSessionTableSQL";

packages/shared/src/server/auth.ts

@@ -1,81 +0,0 @@
-import { compare, hash } from "bcryptjs";
-import { randomUUID } from "crypto";
-import * as crypto from "crypto";
-import type { OAuthConfig, OAuthUserConfig } from "next-auth/providers/oauth";
-
-export function generateSecretKey() {
-  return `sk-lf-${randomUUID()}`;
-}
-
-export function generatePublicKey() {
-  return `pk-lf-${randomUUID()}`;
-}
-
-export function getDisplaySecretKey(secretKey: string) {
-  return secretKey.slice(0, 6) + "..." + secretKey.slice(-4);
-}
-
-export async function hashSecretKey(key: string) {
-  // legacy, uses bcrypt, transformed into hashed key upon first use
-  const hashedKey = await hash(key, 11);
-  return hashedKey;
-}
-
-export async function generateKeySet() {
-  const pk = generatePublicKey();
-  const sk = generateSecretKey();
-  const hashedSk = await hashSecretKey(sk);
-  const displaySk = getDisplaySecretKey(sk);
-
-  return {
-    pk,
-    sk,
-    hashedSk,
-    displaySk,
-  };
-}
-
-export async function verifySecretKey(key: string, hashedKey: string) {
-  const isValid = await compare(key, hashedKey);
-  return isValid;
-}
-
-export function createShaHash(privateKey: string, salt: string): string {
-  const hash = crypto
-    .createHash("sha256")
-    .update(privateKey)
-    .update(crypto.createHash("sha256").update(salt, "utf8").digest("hex"))
-    .digest("hex");
-
-  return hash;
-}
-
-export interface CustomSSOUser extends Record<string, any> {
-  email: string;
-  id: string;
-  name: string;
-  verified: boolean;
-}
-
-export function CustomSSOProvider<P extends CustomSSOUser>(
-  options: OAuthUserConfig<P>
-): OAuthConfig<P> {
-  return {
-    id: "custom",
-    name: "CustomSSOProvider",
-    type: "oauth",
-    wellKnown: `${options.issuer}/.well-known/openid-configuration`,
-    authorization: { params: { scope: "openid email profile" } }, // overridden by options.authorization to be able to set custom scopes, deep merged with this default
-    checks: ["pkce", "state"],
-    idToken: true,
-    profile(profile) {
-      return {
-        id: profile.sub,
-        name: profile.name,
-        email: profile.email,
-        image: null,
-      };
-    },
-    options,
-  };
-}

packages/shared/src/server/auth/apiKeys.ts

@@ -0,0 +1,81 @@
+import { PrismaClient } from "@prisma/client";
+import { compare, hash } from "bcryptjs";
+import { randomUUID } from "crypto";
+import * as crypto from "crypto";
+import { env } from "../../env";
+
+export function getDisplaySecretKey(secretKey: string) {
+  return secretKey.slice(0, 6) + "..." + secretKey.slice(-4);
+}
+
+export async function hashSecretKey(key: string) {
+  // legacy, uses bcrypt, transformed into hashed key upon first use
+  const hashedKey = await hash(key, 11);
+  return hashedKey;
+}
+
+async function generateKeySet() {
+  return {
+    pk: `pk-lf-${randomUUID()}`,
+    sk: `sk-lf-${randomUUID()}`,
+  };
+}
+
+export async function verifySecretKey(key: string, hashedKey: string) {
+  const isValid = await compare(key, hashedKey);
+  return isValid;
+}
+
+export function createShaHash(privateKey: string, salt: string): string {
+  const hash = crypto
+    .createHash("sha256")
+    .update(privateKey)
+    .update(crypto.createHash("sha256").update(salt, "utf8").digest("hex"))
+    .digest("hex");
+
+  return hash;
+}
+
+export async function createAndAddApiKeysToDb(p: {
+  prisma: PrismaClient;
+  projectId: string;
+  note?: string;
+  predefinedKeys?: {
+    secretKey: string;
+    publicKey: string;
+  };
+}) {
+  const salt = env.SALT;
+  if (!salt) {
+    throw new Error("SALT is not set");
+  }
+
+  const { pk, sk } = p.predefinedKeys
+    ? { pk: p.predefinedKeys.publicKey, sk: p.predefinedKeys.secretKey }
+    : await generateKeySet();
+
+  const hashedSk = await hashSecretKey(sk);
+  const displaySk = getDisplaySecretKey(sk);
+
+  const hashFromProvidedKey = createShaHash(sk, salt);
+
+  const apiKey = await p.prisma.apiKey.create({
+    data: {
+      projectId: p.projectId,
+      publicKey: pk,
+      hashedSecretKey: hashedSk,
+      displaySecretKey: displaySk,
+      fastHashedSecretKey: hashFromProvidedKey,
+      note: p.note,
+    },
+  });
+
+  return {
+    id: apiKey.id,
+    createdAt: apiKey.createdAt,
+    note: apiKey.note,
+    publicKey: apiKey.publicKey,
+    secretKey: sk,
+    displaySecretKey: displaySk,
+  };
+}

packages/shared/src/server/auth/customSsoProvider.ts

@@ -0,0 +1,31 @@
+import type { OAuthConfig, OAuthUserConfig } from "next-auth/providers/oauth";
+
+interface CustomSSOUser extends Record<string, any> {
+  email: string;
+  id: string;
+  name: string;
+  verified: boolean;
+}
+
+export function CustomSSOProvider<P extends CustomSSOUser>(
+  options: OAuthUserConfig<P>
+): OAuthConfig<P> {
+  return {
+    id: "custom",
+    name: "CustomSSOProvider",
+    type: "oauth",
+    wellKnown: `${options.issuer}/.well-known/openid-configuration`,
+    authorization: { params: { scope: "openid email profile" } }, // overridden by options.authorization to be able to set custom scopes, deep merged with this default
+    checks: ["pkce", "state"],
+    idToken: true,
+    profile(profile) {
+      return {
+        id: profile.sub,
+        name: profile.name,
+        email: profile.email,
+        image: null,
+      };
+    },
+    options,
+  };
+}

packages/shared/src/server/auth/types.ts

@@ -0,0 +1,46 @@
+import z from "zod";
+import { Plan, plans } from "../../features/entitlements/plans";
+import { CloudConfigRateLimit } from "../../interfaces/rate-limits";
+
+export const OrgEnrichedApiKey = z.object({
+  id: z.string(),
+  note: z.string().nullable(),
+  publicKey: z.string(),
+  displaySecretKey: z.string(),
+  createdAt: z.string().datetime().nullable(),
+  lastUsedAt: z.string().datetime().nullable(),
+  expiresAt: z.string().datetime().nullable(),
+  projectId: z.string(),
+  fastHashedSecretKey: z.string(),
+  hashedSecretKey: z.string(),
+  orgId: z.string(),
+  plan: z.enum(plans as unknown as [string, ...string[]]),
+  rateLimitOverrides: CloudConfigRateLimit.nullish(),
+});
+
+export const API_KEY_NON_EXISTENT = "api-key-non-existent";
+
+export const CachedApiKey = z.union([
+  OrgEnrichedApiKey,
+  z.literal(API_KEY_NON_EXISTENT),
+]);
+
+export type AuthHeaderVerificationResult =
+  | AuthHeaderValidVerificationResult
+  | {
+      validKey: false;
+      error: string;
+    };
+
+export type AuthHeaderValidVerificationResult = {
+  validKey: true;
+  scope: ApiAccessScope;
+};
+
+export type ApiAccessScope = {
+  projectId: string;
+  accessLevel: "all" | "scores";
+  orgId: string;
+  plan: Plan;
+  rateLimitOverrides: z.infer<typeof CloudConfigRateLimit>;
+};

packages/shared/src/server/clickhouse.ts

@@ -0,0 +1,16 @@
+import { createClient } from "@clickhouse/client";
+
+import { env } from "../env";
+
+export type ClickhouseClientType = ReturnType<typeof createClient>;
+
+export const clickhouseClient = createClient({
+  url: env.CLICKHOUSE_URL,
+  username: env.CLICKHOUSE_USER,
+  password: env.CLICKHOUSE_PASSWORD,
+  database: "default",
+  clickhouse_settings: {
+    async_insert: 1,
+    wait_for_async_insert: 1, // if disabled, we won't get errors from clickhouse
+  },
+});

packages/shared/src/server/definitions.ts

@@ -0,0 +1,167 @@
+import z from "zod";
+
+export const clickhouseStringDateSchema = z
+  .string()
+  // clickhouse stores UTC like '2024-05-23 18:33:41.602000'
+  // we need to convert it to '2024-05-23T18:33:41.602000Z'
+  .transform((str) => str.replace(" ", "T") + "Z")
+  .pipe(z.string().datetime());
+
+export const observationRecordBaseSchema = z.object({
+  id: z.string(),
+  trace_id: z.string().nullish(),
+  project_id: z.string(),
+  type: z.string(),
+  parent_observation_id: z.string().nullish(),
+  name: z.string().nullish(),
+  metadata: z.record(z.string()),
+  level: z.string().nullish(),
+  status_message: z.string().nullish(),
+  version: z.string().nullish(),
+  input: z.string().nullish(),
+  output: z.string().nullish(),
+  provided_model_name: z.string().nullish(),
+  internal_model_id: z.string().nullish(),
+  model_parameters: z.string().nullish(),
+  unit: z.string().nullish(),
+  input_usage_units: z.number().nullish(),
+  output_usage_units: z.number().nullish(),
+  total_usage_units: z.number().nullish(),
+  input_cost: z.number().nullish(),
+  output_cost: z.number().nullish(),
+  total_cost: z.number().nullish(),
+  provided_input_usage_units: z.number().nullish(),
+  provided_output_usage_units: z.number().nullish(),
+  provided_total_usage_units: z.number().nullish(),
+  provided_input_cost: z.number().nullish(),
+  provided_output_cost: z.number().nullish(),
+  provided_total_cost: z.number().nullish(),
+  prompt_id: z.string().nullish(),
+  prompt_name: z.string().nullish(),
+  prompt_version: z.number().nullish(),
+});
+export type ObservationRecordBaseType = z.infer<
+  typeof observationRecordBaseSchema
+>;
+
+export const observationRecordReadSchema = observationRecordBaseSchema.extend({
+  created_at: clickhouseStringDateSchema,
+  updated_at: clickhouseStringDateSchema,
+  start_time: clickhouseStringDateSchema,
+  end_time: clickhouseStringDateSchema.nullish(),
+  completion_start_time: clickhouseStringDateSchema.nullish(),
+});
+export type ObservationRecordReadType = z.infer<
+  typeof observationRecordReadSchema
+>;
+
+export const observationRecordInsertSchema = observationRecordBaseSchema.extend(
+  {
+    created_at: z.number(),
+    updated_at: z.number(),
+    start_time: z.number(),
+    end_time: z.number().nullish(),
+    completion_start_time: z.number().nullish(),
+  }
+);
+export type ObservationRecordInsertType = z.infer<
+  typeof observationRecordInsertSchema
+>;
+
+export const traceRecordBaseSchema = z.object({
+  id: z.string(),
+  name: z.string().nullish(),
+  user_id: z.string().nullish(),
+  metadata: z.record(z.string()),
+  release: z.string().nullish(),
+  version: z.string().nullish(),
+  project_id: z.string(),
+  public: z.boolean(),
+  bookmarked: z.boolean(),
+  tags: z.array(z.string()),
+  input: z.string().nullish(),
+  output: z.string().nullish(),
+  session_id: z.string().nullish(),
+});
+export type TraceRecordBaseType = z.infer<typeof traceRecordBaseSchema>;
+
+export const traceRecordReadSchema = traceRecordBaseSchema.extend({
+  timestamp: clickhouseStringDateSchema,
+  created_at: clickhouseStringDateSchema,
+  updated_at: clickhouseStringDateSchema,
+});
+export type TraceRecordReadType = z.infer<typeof traceRecordReadSchema>;
+
+export const traceRecordInsertSchema = traceRecordBaseSchema.extend({
+  timestamp: z.number(),
+  created_at: z.number(),
+  updated_at: z.number(),
+});
+export type TraceRecordInsertType = z.infer<typeof traceRecordInsertSchema>;
+
+export const scoreRecordBaseSchema = z.object({
+  id: z.string(),
+  project_id: z.string(),
+  trace_id: z.string(),
+  observation_id: z.string().nullish(),
+  name: z.string().nullish(),
+  value: z.union([z.number(), z.string()]).nullish(),
+  source: z.string(),
+  comment: z.string().nullish(),
+  author_user_id: z.string().nullish(),
+  config_id: z.string().nullish(),
+  data_type: z.enum(["NUMERIC", "CATEGORICAL", "BOOLEAN"]).nullish(),
+  string_value: z.string().nullish(),
+});
+export type ScoreRecordBaseType = z.infer<typeof scoreRecordBaseSchema>;
+
+export const scoreRecordReadSchema = scoreRecordBaseSchema.extend({
+  created_at: clickhouseStringDateSchema,
+  updated_at: clickhouseStringDateSchema,
+  timestamp: clickhouseStringDateSchema,
+});
+export type ScoreRecordReadType = z.infer<typeof scoreRecordReadSchema>;
+
+export const scoreRecordInsertSchema = scoreRecordBaseSchema.extend({
+  created_at: z.number(),
+  updated_at: z.number(),
+  timestamp: z.number(),
+});
+export type ScoreRecordInsertType = z.infer<typeof scoreRecordInsertSchema>;
+
+export const convertTraceReadToInsert = (
+  record: TraceRecordReadType
+): TraceRecordInsertType => {
+  return {
+    ...record,
+    created_at: new Date(record.created_at).getTime(),
+    updated_at: new Date(record.created_at).getTime(),
+    timestamp: new Date(record.timestamp).getTime(),
+  };
+};
+
+export const convertObservationReadToInsert = (
+  record: ObservationRecordReadType
+): ObservationRecordInsertType => {
+  return {
+    ...record,
+    created_at: new Date(record.created_at).getTime(),
+    updated_at: new Date(record.created_at).getTime(),
+    start_time: new Date(record.start_time).getTime(),
+    end_time: record.end_time ? new Date(record.end_time).getTime() : undefined,
+    completion_start_time: record.completion_start_time
+      ? new Date(record.completion_start_time).getTime()
+      : undefined,
+  };
+};
+
+export const convertScoreReadToInsert = (
+  record: ScoreRecordReadType
+): ScoreRecordInsertType => {
+  return {
+    ...record,
+    created_at: new Date(record.created_at).getTime(),
+    updated_at: new Date(record.updated_at).getTime(),
+    timestamp: new Date(record.timestamp).getTime(),
+  };
+};

packages/shared/src/server/filterToPrisma.ts

@@ -1,7 +1,9 @@
 import { Prisma } from "@prisma/client";
-import { ColumnDefinition, type TableNames } from "./tableDefinitions";
-import { FilterState } from "./types";
-import { filterOperators } from "./interfaces/filters";
+import { ColumnDefinition, type TableNames } from "../tableDefinitions";
+import { FilterState } from "../types";
+import { filterOperators, timeFilter } from "../interfaces/filters";
+import { z } from "zod";
+import { logger } from "./index";
 
 const operatorReplacements = {
   "any of": "IN",
@@ -24,7 +26,7 @@ const arrayOperatorReplacements = {
 export function tableColumnsToSqlFilterAndPrefix(
   filters: FilterState,
   tableColumns: ColumnDefinition[],
-  table: TableNames
+  table: TableNames,
 ): Prisma.Sql {
   const sql = tableColumnsToSqlFilter(filters, tableColumns, table);
   if (sql === Prisma.empty) {
@@ -40,17 +42,17 @@ export function tableColumnsToSqlFilterAndPrefix(
 export function tableColumnsToSqlFilter(
   filters: FilterState,
   tableColumns: ColumnDefinition[],
-  table: TableNames
+  table: TableNames,
 ): Prisma.Sql {
   const internalFilters = filters.map((filter) => {
     // Get column definition to map column to internal name, e.g. "t.id"
     const col = tableColumns.find(
       (c) =>
         // TODO: Only use id instead of name
-        c.name === filter.column || c.id === filter.column
+        c.name === filter.column || c.id === filter.column,
     );
     if (!col) {
-      console.error("Invalid filter column", filter.column);
+      logger.error("Invalid filter column", filter.column);
       throw new Error("Invalid filter column: " + filter.column);
     }
     const colPrisma = Prisma.raw(col.internal);
@@ -69,13 +71,13 @@ export function tableColumnsToSqlFilter(
         ? Prisma.raw(
             arrayOperatorReplacements[
               filter.operator as keyof typeof arrayOperatorReplacements
-            ]
+            ],
           )
         : filter.operator in operatorReplacements
           ? Prisma.raw(
               operatorReplacements[
                 filter.operator as keyof typeof operatorReplacements
-              ]
+              ],
             )
           : Prisma.raw(filter.operator); //checked by zod
 
@@ -95,13 +97,13 @@ export function tableColumnsToSqlFilter(
         break;
       case "stringOptions":
         valuePrisma = Prisma.sql`(${Prisma.join(
-          filter.value.map((v) => Prisma.sql`${v}`)
+          filter.value.map((v) => Prisma.sql`${v}`),
         )})`;
         break;
       case "arrayOptions":
         valuePrisma = Prisma.sql`ARRAY[${Prisma.join(
           filter.value.map((v) => Prisma.sql`${v}`),
-          ", "
+          ", ",
         )}] `;
         break;
 
@@ -121,12 +123,12 @@ export function tableColumnsToSqlFilter(
       filter.type === "string" || filter.type === "stringObject"
         ? [
             ["contains", "does not contain", "ends with"].includes(
-              filter.operator
+              filter.operator,
             )
               ? Prisma.raw("'%' || ")
               : Prisma.empty,
             ["contains", "does not contain", "starts with"].includes(
-              filter.operator
+              filter.operator,
             )
               ? Prisma.raw(" || '%'")
               : Prisma.empty,
@@ -150,7 +152,7 @@ export function tableColumnsToSqlFilter(
 
 const castValueToPostgresTypes = (
   column: ColumnDefinition,
-  table: TableNames
+  table: TableNames,
 ) => {
   return column.name === "type" &&
     (table === "observations" ||
@@ -166,7 +168,7 @@ const dateOperators = filterOperators["datetime"];
 export const datetimeFilterToPrismaSql = (
   safeColumn: string,
   operator: (typeof dateOperators)[number],
-  value: Date
+  value: Date,
 ) => {
   if (!dateOperators.includes(operator)) {
     throw new Error("Invalid operator: " + operator);
@@ -176,6 +178,22 @@ export const datetimeFilterToPrismaSql = (
   }
 
   return Prisma.sql`AND ${Prisma.raw(safeColumn)} ${Prisma.raw(
-    operator
+    operator,
   )} ${value}::timestamp with time zone at time zone 'UTC'`;
 };
+
+export const datetimeFilterToPrisma = (
+  timestampFilter: z.infer<typeof timeFilter>,
+) => {
+  const prismaTimestampFilter =
+    timestampFilter.operator === ">="
+      ? { gte: timestampFilter.value }
+      : timestampFilter.operator === ">"
+        ? { gt: timestampFilter.value }
+        : timestampFilter.operator === "<="
+          ? { lte: timestampFilter.value }
+          : timestampFilter.operator === "<"
+            ? { lt: timestampFilter.value }
+            : {};
+  return prismaTimestampFilter;
+};

packages/shared/src/server/index.ts

@@ -1,9 +1,32 @@
 export * from "./services/S3StorageService";
-export * from "./services/email/projectInvitation/sendProjectInvitationEmail";
+export * from "./services/email/organizationInvitation/sendMembershipInvitationEmail";
 export * from "./services/email/batchExportSuccess/sendBatchExportSuccessEmail";
 export * from "./services/email/passwordReset/sendResetPasswordVerificationRequest";
-export * from "./auth";
+export * from "./services/PromptService";
+export * from "./auth/apiKeys";
+export * from "./auth/customSsoProvider";
 export * from "./llm/fetchLLMCompletion";
 export * from "./llm/types";
 export * from "./utils/DatabaseReadStream";
 export * from "./utils/transforms";
+export * from "./clickhouse";
+export * from "../server/definitions";
+export * from "../server/ingestion/IngestionUtils";
+export * from "../server/ingestion/types";
+export * from "../server/ingestion/model-match";
+export * from "../server/ingestion/types";
+export * from "../server/ingestion/validateAndInflateScore";
+export * from "./redis/redis";
+export * from "./redis/trace-upsert";
+export * from "./redis/batch-export";
+export * from "./redis/ingestionFlushQueue";
+export * from "./redis/legacy-ingestion";
+export * from "./auth/types";
+export * from "./ingestion/legacy/index";
+export * from "./queues";
+export * from "./ingestion/legacy/EventProcessor";
+export * from "./orderByToPrisma";
+export * from "./filterToPrisma";
+export * from "./instrumentation";
+export * from "./logger";
+export * from "./queries";

packages/shared/src/server/ingestion/IngestionUtils.ts

@@ -0,0 +1,86 @@
+import { eventTypes, IngestionEventType } from "./types";
+
+const reservedCharsEscapeMap = [
+  { reserved: ":", escape: "|%|" },
+  { reserved: "_", escape: "|#|" },
+];
+
+export enum ClickhouseEntityType {
+  Trace = "trace",
+  Score = "score",
+  Observation = "observation",
+  SDK_LOG = "sdk-log",
+}
+
+export class IngestionUtils {
+  public static getBufferKey(flushKey: string): string {
+    const { projectId, eventType, entityId } =
+      IngestionUtils.parseFlushKey(flushKey);
+
+    const sanitizedEntityId = IngestionUtils.escapeReservedChars(entityId);
+
+    return (
+      "ingestionBuffer:" + `${projectId}_${eventType}_${sanitizedEntityId}`
+    );
+  }
+
+  public static getFlushKey(params: {
+    projectId: string;
+    eventType: ClickhouseEntityType;
+    entityId: string;
+    batchTimestamp: string;
+  }): string {
+    const { projectId, eventType, entityId, batchTimestamp } = params;
+    const sanitizedEntityId = IngestionUtils.escapeReservedChars(entityId);
+
+    return `${projectId}_${eventType}_${sanitizedEntityId}_${batchTimestamp}`;
+  }
+
+  public static parseFlushKey(projectEntityKey: string) {
+    const split = projectEntityKey.split("_");
+
+    if (split.length < 3) {
+      throw new Error(
+        `Invalid project entity key format ${projectEntityKey}, expected 3 or 4 parts`
+      );
+    }
+
+    const [projectId, eventType, escapedEntityId, batchTimestamp] = split;
+    const entityId = IngestionUtils.unescapeReservedChars(escapedEntityId);
+
+    return { projectId, eventType, entityId, batchTimestamp };
+  }
+
+  private static escapeReservedChars(string: string): string {
+    return reservedCharsEscapeMap.reduce(
+      (acc, { reserved, escape }) => acc.replaceAll(reserved, escape),
+      string
+    );
+  }
+
+  private static unescapeReservedChars(escapedString: string): string {
+    return reservedCharsEscapeMap.reduce(
+      (acc, { reserved, escape }) => acc.replaceAll(escape, reserved),
+      escapedString
+    );
+  }
+
+  public static getEventType(event: IngestionEventType): ClickhouseEntityType {
+    switch (event.type) {
+      case eventTypes.TRACE_CREATE:
+        return ClickhouseEntityType.Trace;
+      case eventTypes.OBSERVATION_CREATE:
+      case eventTypes.OBSERVATION_UPDATE:
+      case eventTypes.EVENT_CREATE:
+      case eventTypes.SPAN_CREATE:
+      case eventTypes.SPAN_UPDATE:
+      case eventTypes.GENERATION_CREATE:
+      case eventTypes.GENERATION_UPDATE:
+        return ClickhouseEntityType.Observation;
+      case eventTypes.SCORE_CREATE:
+        return ClickhouseEntityType.Score;
+      case eventTypes.SDK_LOG:
+        return ClickhouseEntityType.SDK_LOG;
+    }
+  }
+}

packages/shared/src/server/ingestion/legacy/EventProcessor.ts

@@ -1,130 +1,53 @@
-import { tokenCount } from "@/src/features/ingest/lib/usage";
-import { type ApiAccessScope } from "@/src/features/public-api/server/types";
-import {
-  type legacyObservationCreateEvent,
-  eventTypes,
-  type scoreEvent,
-  type eventCreateEvent,
-  type spanCreateEvent,
-  type generationCreateEvent,
-  type spanUpdateEvent,
-  type generationUpdateEvent,
-  type legacyObservationUpdateEvent,
-  type sdkLogEvent,
-  type traceEvent,
-  LangfuseNotFoundError,
-  InvalidRequestError,
-} from "@langfuse/shared";
-import { ScoreDataType, prisma } from "@langfuse/shared/src/db";
-import { ResourceNotFoundError } from "@/src/utils/exceptions";
-import { mergeJson } from "@langfuse/shared";
-import {
-  type Trace,
-  type Observation,
-  type Score,
-  Prisma,
-  type Model,
-} from "@langfuse/shared/src/db";
 import { v4 } from "uuid";
 import { type z } from "zod";
-import { jsonSchema } from "@langfuse/shared";
-import { ForbiddenError } from "@langfuse/shared";
-import { instrument } from "@/src/utils/instrumentation";
 import Decimal from "decimal.js";
+import { findModel } from "../model-match";
 import {
-  ScoreBodyWithoutConfig,
-  ScorePropsAgainstConfig,
-} from "@/src/features/public-api/types/scores";
-import {
-  validateDbScoreConfigSafe,
-  type ValidatedScoreConfig,
-} from "@/src/features/public-api/types/score-configs";
+  ObservationEvent,
+  eventTypes,
+  legacyObservationCreateEvent,
+  generationCreateEvent,
+  traceEvent,
+  scoreEvent,
+  sdkLogEvent,
+} from "../types";
+import { validateAndInflateScore } from "../validateAndInflateScore";
+import { Trace, Observation, Score, Prisma, Model } from "@prisma/client";
+import { ForbiddenError, LangfuseNotFoundError } from "../../../errors";
+import { mergeJson } from "../../../utils/json";
+import { jsonSchema } from "../../../utils/zod";
+import { prisma } from "../../../db";
+import { LegacyIngestionAccessScope } from ".";
+import { logger } from "../../logger";
 
 export interface EventProcessor {
+  auth(apiScope: LegacyIngestionAccessScope): void;
+
   process(
-    apiScope: ApiAccessScope,
+    apiScope: LegacyIngestionAccessScope,
   ): Promise<Trace | Observation | Score> | undefined;
 }
 
-export async function findModel(p: {
-  event: {
-    projectId: string;
-    model?: string;
-    unit?: string;
-    startTime?: Date;
-  };
-  existingDbObservation?: Observation;
-}): Promise<Model | null> {
-  const { event, existingDbObservation } = p;
-  // either get the model from the existing observation
-  // or match pattern on the user provided model name
-  const modelCondition = event.model
-    ? Prisma.sql`AND ${event.model} ~ match_pattern`
-    : existingDbObservation?.internalModel
-      ? Prisma.sql`AND model_name = ${existingDbObservation.internalModel}`
-      : undefined;
-  if (!modelCondition) return null;
-
-  // unit based on the current event or the existing observation, both can be undefined
-  const mergedUnit = event.unit ?? existingDbObservation?.unit;
-
-  const unitCondition = mergedUnit
-    ? Prisma.sql`AND unit = ${mergedUnit}`
-    : Prisma.empty;
-
-  const sql = Prisma.sql`
-    SELECT
-      id,
-      created_at AS "createdAt",
-      updated_at AS "updatedAt",
-      project_id AS "projectId",
-      model_name AS "modelName",
-      match_pattern AS "matchPattern",
-      start_date AS "startDate",
-      input_price AS "inputPrice",
-      output_price AS "outputPrice",
-      total_price AS "totalPrice",
-      unit, 
-      tokenizer_id AS "tokenizerId",
-      tokenizer_config AS "tokenizerConfig"
-    FROM
-      models
-    WHERE (project_id = ${event.projectId}
-      OR project_id IS NULL)
-    ${modelCondition}
-    ${unitCondition}
-    AND (start_date IS NULL OR start_date <= ${
-      event.startTime ? new Date(event.startTime) : new Date()
-    }::timestamp with time zone at time zone 'UTC')
-    ORDER BY
-      project_id ASC,
-      start_date DESC NULLS LAST
-    LIMIT 1
-  `;
-
-  const foundModels = await prisma.$queryRaw<Array<Model>>(sql);
-
-  return foundModels[0] ?? null;
-}
-
-type ObservationEvent =
-  | z.infer<typeof legacyObservationCreateEvent>
-  | z.infer<typeof legacyObservationUpdateEvent>
-  | z.infer<typeof eventCreateEvent>
-  | z.infer<typeof spanCreateEvent>
-  | z.infer<typeof spanUpdateEvent>
-  | z.infer<typeof generationCreateEvent>
-  | z.infer<typeof generationUpdateEvent>;
-
 export class ObservationProcessor implements EventProcessor {
   event: ObservationEvent;
+  calculateTokenDelegate: (p: {
+    model: Model;
+    text: unknown;
+  }) => number | undefined;
 
-  constructor(event: ObservationEvent) {
+  constructor(
+    event: ObservationEvent,
+    calculateTokenDelegate: (p: {
+      model: Model;
+      text: unknown;
+    }) => number | undefined,
+  ) {
     this.event = event;
+    this.calculateTokenDelegate = calculateTokenDelegate;
   }
 
   async convertToObservation(
-    apiScope: ApiAccessScope,
+    apiScope: LegacyIngestionAccessScope,
     existingObservation: Observation | null,
   ): Promise<{
     id: string;
@@ -154,7 +77,9 @@ export class ObservationProcessor implements EventProcessor {
       this.event.type === eventTypes.OBSERVATION_UPDATE &&
       !existingObservation
     ) {
-      throw new ResourceNotFoundError(this.event.id, "Observation not found");
+      throw new LangfuseNotFoundError(
+        `Observation with id ${this.event.id} not found`,
+      );
     }
 
     // find matching model definition based on event and existing observation in db
@@ -197,6 +122,7 @@ export class ObservationProcessor implements EventProcessor {
       "usage" in this.event.body
         ? this.calculateTokenCounts(
             this.event.body,
+            this.calculateTokenDelegate,
             internalModel ?? undefined,
             existingObservation ?? undefined,
           )
@@ -262,8 +188,9 @@ export class ObservationProcessor implements EventProcessor {
         : undefined;
 
     // Only null if promptName and promptVersion are set but prompt is not found
-    if (prompt === null)
-      console.warn("Prompt not found for observation", this.event.body);
+    if (prompt === null) {
+      logger.warn("Prompt not found for observation", this.event.body);
+    }
 
     const observationId = this.event.body.id ?? v4();
 
@@ -388,34 +315,34 @@ export class ObservationProcessor implements EventProcessor {
     body:
       | z.infer<typeof legacyObservationCreateEvent>["body"]
       | z.infer<typeof generationCreateEvent>["body"],
+    calculateTokenDelegate: (p: {
+      model: Model;
+      text: unknown;
+    }) => number | undefined,
     model?: Model,
     existingObservation?: Observation,
   ) {
-    return instrument({ name: "calculate-tokens" }, () => {
-      const newPromptTokens =
-        body.usage?.input ??
-        ((body.input || existingObservation?.input) &&
-        model &&
-        model.tokenizerId
-          ? tokenCount({
-              model: model,
-              text: body.input ?? existingObservation?.input,
-            })
-          : undefined);
+    const newPromptTokens =
+      body.usage?.input ??
+      ((body.input || existingObservation?.input) && model && model.tokenizerId
+        ? calculateTokenDelegate({
+            model: model,
+            text: body.input ?? existingObservation?.input,
+          })
+        : undefined);
 
-      const newCompletionTokens =
-        body.usage?.output ??
-        ((body.output || existingObservation?.output) &&
-        model &&
-        model.tokenizerId
-          ? tokenCount({
-              model: model,
-              text: body.output ?? existingObservation?.output,
-            })
-          : undefined);
+    const newCompletionTokens =
+      body.usage?.output ??
+      ((body.output || existingObservation?.output) &&
+      model &&
+      model.tokenizerId
+        ? calculateTokenDelegate({
+            model: model,
+            text: body.output ?? existingObservation?.output,
+          })
+        : undefined);
 
-      return [newPromptTokens, newCompletionTokens];
-    });
+    return [newPromptTokens, newCompletionTokens];
   }
 
   static calculateTokenCosts(
@@ -473,9 +400,13 @@ export class ObservationProcessor implements EventProcessor {
     };
   }
 
-  async process(apiScope: ApiAccessScope): Promise<Observation> {
+  auth(apiScope: LegacyIngestionAccessScope): void {
     if (apiScope.accessLevel !== "all")
       throw new ForbiddenError("Access denied for observation creation");
+  }
+
+  async process(apiScope: LegacyIngestionAccessScope): Promise<Observation> {
+    this.auth(apiScope);
 
     const existingObservation = this.event.body.id
       ? await prisma.observation.findFirst({
@@ -512,23 +443,22 @@ export class TraceProcessor implements EventProcessor {
     this.event = event;
   }
 
+  auth(apiScope: LegacyIngestionAccessScope): void {
+    if (apiScope.accessLevel !== "all")
+      throw new ForbiddenError("Access denied for trace creation");
+  }
+
   async process(
-    apiScope: ApiAccessScope,
+    apiScope: LegacyIngestionAccessScope,
   ): Promise<Trace | Observation | Score> {
     const { body } = this.event;
 
-    if (apiScope.accessLevel !== "all")
-      throw new ForbiddenError(
-        `Access denied for trace creation, ${apiScope.accessLevel}`,
-      );
+    this.auth(apiScope);
 
     const internalId = body.id ?? v4();
 
-    console.log(
-      "Trying to create trace, project ",
-      apiScope.projectId,
-      ", id:",
-      internalId,
+    logger.info(
+      `Trying to create trace, project ${apiScope.projectId}, id: ${internalId}`,
     );
 
     const existingTrace = await prisma.trace.findFirst({
@@ -622,154 +552,19 @@ export class ScoreProcessor implements EventProcessor {
     this.event = event;
   }
 
-  static inferDataType(value: string | number): ScoreDataType {
-    return typeof value === "number"
-      ? ScoreDataType.NUMERIC
-      : ScoreDataType.CATEGORICAL;
-  }
-
-  static mapStringValueToNumericValue(
-    config: ValidatedScoreConfig,
-    label: string,
-  ): number | null {
-    return (
-      config.categories?.find((category) => category.label === label)?.value ??
-      null
-    );
-  }
-
-  static inflateScoreBody(
-    body: any,
-    id: string,
-    projectId: string,
-    config?: ValidatedScoreConfig,
-  ): Score {
-    const relevantDataType = config?.dataType ?? body.dataType;
-    const scoreProps = { ...body, id, projectId, source: "API" };
-
-    if (typeof body.value === "number") {
-      if (relevantDataType && relevantDataType === ScoreDataType.BOOLEAN) {
-        return {
-          ...scoreProps,
-          value: body.value,
-          stringValue: body.value === 1 ? "True" : "False",
-          dataType: ScoreDataType.BOOLEAN,
-        };
-      }
-
-      return {
-        ...scoreProps,
-        value: body.value,
-        dataType: ScoreDataType.NUMERIC,
-      };
-    }
-    return {
-      ...scoreProps,
-      value: config
-        ? ScoreProcessor.mapStringValueToNumericValue(config, body.value)
-        : null,
-      stringValue: body.value,
-      dataType: ScoreDataType.CATEGORICAL,
-    };
-  }
-
-  validateConfigAgainstBody(body: any, config: ValidatedScoreConfig): void {
-    const { maxValue, minValue, categories, dataType: configDataType } = config;
-    if (body.dataType && body.dataType !== configDataType) {
-      throw new InvalidRequestError(
-        `Data type mismatch based on config: expected ${configDataType}, got ${body.dataType}`,
-      );
-    }
-
-    if (config.isArchived) {
-      throw new InvalidRequestError(
-        "Config is archived and cannot be used to create new scores. Please restore the config first.",
-      );
-    }
-
-    if (config.name !== body.name) {
-      throw new InvalidRequestError(
-        `Name mismatch based on config: expected ${config.name}, got ${body.name}`,
-      );
-    }
-
-    const relevantDataType = configDataType ?? body.dataType;
-
-    const dataTypeValidation = ScoreBodyWithoutConfig.safeParse({
-      ...body,
-      dataType: relevantDataType,
-    });
-    if (!dataTypeValidation.success) {
-      throw new InvalidRequestError(
-        `Ingested score body not valid against provided config data type.`,
-      );
-    }
-
-    const rangeValidation = ScorePropsAgainstConfig.safeParse({
-      value: body.value,
-      dataType: relevantDataType,
-      ...(maxValue !== null && maxValue !== undefined && { maxValue }),
-      ...(minValue !== null && minValue !== undefined && { minValue }),
-      ...(categories && { categories }),
-    });
-    if (!rangeValidation.success) {
-      const errorDetails = rangeValidation.error.errors
-        .map((error) => `${error.path.join(".")} - ${error.message}`)
-        .join(", ");
-      throw new InvalidRequestError(
-        `Ingested score body not valid against provided config: ${errorDetails}`,
-      );
-    }
-  }
-
-  async validateAndInflate(
-    body: any,
-    id: string,
-    projectId: string,
-  ): Promise<Score> {
-    if (body.configId) {
-      const config = await prisma.scoreConfig.findFirst({
-        where: {
-          projectId,
-          id: body.configId,
-        },
-      });
-
-      if (!config || !validateDbScoreConfigSafe(config).success)
-        throw new LangfuseNotFoundError(
-          "The configId you provided does not match a valid config in this project",
-        );
-
-      this.validateConfigAgainstBody(body, config as ValidatedScoreConfig);
-      return ScoreProcessor.inflateScoreBody(
-        body,
-        id,
-        projectId,
-        config as ValidatedScoreConfig,
-      );
-    } else {
-      const validation = ScoreBodyWithoutConfig.safeParse({
-        ...body,
-        dataType: body.dataType ?? ScoreProcessor.inferDataType(body.value),
-      });
-      if (!validation.success) {
-        throw new InvalidRequestError(
-          `Ingested score value type not valid against provided data type. Provide numeric values for numeric and boolean scores, and string values for categorical scores.`,
-        );
-      }
-      return ScoreProcessor.inflateScoreBody(body, id, projectId);
-    }
-  }
-
-  async process(
-    apiScope: ApiAccessScope,
-  ): Promise<Trace | Observation | Score> {
-    const { body } = this.event;
-
+  auth(apiScope: LegacyIngestionAccessScope) {
     if (apiScope.accessLevel !== "scores" && apiScope.accessLevel !== "all")
       throw new ForbiddenError(
         `Access denied for score creation, ${apiScope.accessLevel}`,
       );
+  }
+
+  async process(
+    apiScope: LegacyIngestionAccessScope,
+  ): Promise<Trace | Observation | Score> {
+    const { body } = this.event;
+
+    this.auth(apiScope);
 
     const id = body.id ?? v4();
 
@@ -787,11 +582,11 @@ export class ScoreProcessor implements EventProcessor {
       );
     }
 
-    const validatedScore = await this.validateAndInflate(
+    const validatedScore = await validateAndInflateScore({
       body,
-      id,
-      apiScope.projectId,
-    );
+      scoreId: id,
+      projectId: apiScope.projectId,
+    });
 
     return await prisma.score.upsert({
       where: {
@@ -817,9 +612,13 @@ export class SdkLogProcessor implements EventProcessor {
     this.event = event;
   }
 
-  process(apiScope: ApiAccessScope) {
+  auth(apiScope: LegacyIngestionAccessScope) {
+    return;
+  }
+
+  process() {
     try {
-      console.log("SDK Log", this.event);
+      logger.info("SDK Log", this.event);
       return undefined;
     } catch (error) {
       return undefined;

packages/shared/src/server/ingestion/legacy/enqueueIngestionEvents.ts

@@ -0,0 +1,72 @@
+import { type Redis } from "ioredis";
+import { QueueJobs } from "../../queues";
+import {
+  getIngestionFlushQueue,
+  IngestionFlushQueue,
+} from "../../redis/ingestionFlushQueue";
+import { IngestionUtils } from "../IngestionUtils";
+import { IngestionEventType } from "../types";
+import { redis } from "../../redis/redis";
+import { env } from "../../../env";
+import { logger } from "../../logger";
+
+export async function enqueueIngestionEvents(
+  projectId: string,
+  events: IngestionEventType[],
+) {
+  const ingestionFlushQueue = getIngestionFlushQueue();
+
+  if (!ingestionFlushQueue) {
+    throw Error("IngestionFlushQueue not initialized");
+  }
+  if (!redis) throw Error("Redis connection not available");
+
+  const queuedEventPromises: Promise<void>[] = [];
+  const batchTimestamp = Date.now().toString();
+
+  // Use for loop as TS does not narrow redis type in map function
+  for (const event of events) {
+    queuedEventPromises.push(
+      enqueueSingleIngestionEvent(
+        projectId,
+        event,
+        redis,
+        ingestionFlushQueue,
+        batchTimestamp,
+      ),
+    );
+  }
+
+  await Promise.all(queuedEventPromises);
+}
+
+async function enqueueSingleIngestionEvent(
+  projectId: string,
+  event: IngestionEventType,
+  redis: Redis,
+  ingestionFlushQueue: IngestionFlushQueue,
+  batchTimestamp: string,
+): Promise<void> {
+  if (!("id" in event.body && event.body.id)) {
+    logger.warn(
+      `Received ingestion event without id: ${JSON.stringify(event)}`,
+    );
+
+    return;
+  }
+
+  const flushKey = IngestionUtils.getFlushKey({
+    entityId: event.body.id,
+    eventType: IngestionUtils.getEventType(event),
+    projectId,
+    batchTimestamp,
+  });
+  const bufferKey = IngestionUtils.getBufferKey(flushKey);
+  const serializedEventData = JSON.stringify({ ...event, projectId });
+
+  await redis.lpush(bufferKey, serializedEventData);
+  await redis.expire(bufferKey, env.LANGFUSE_INGESTION_BUFFER_TTL_SECONDS);
+  await ingestionFlushQueue.add(QueueJobs.FlushIngestionEntity, null, {
+    jobId: flushKey,
+  });
+}

packages/shared/src/server/ingestion/legacy/index.ts

@@ -0,0 +1,268 @@
+import { env } from "node:process";
+import z from "zod";
+import { ForbiddenError, UnauthorizedError } from "../../../errors";
+import { eventTypes, ingestionApiSchema, ingestionEvent } from "../types";
+import {
+  EventProcessor,
+  TraceProcessor,
+  ObservationProcessor,
+  ScoreProcessor,
+  SdkLogProcessor,
+} from "./EventProcessor";
+import { EventBodyType, EventName, TraceUpsertEventType } from "../../queues";
+import {
+  convertTraceUpsertEventsToRedisEvents,
+  getTraceUpsertQueue,
+} from "../../redis/trace-upsert";
+import { ApiAccessScope } from "../../auth/types";
+import { redis } from "../../redis/redis";
+import { backOff } from "exponential-backoff";
+import { Model } from "../../..";
+import { enqueueIngestionEvents } from "./enqueueIngestionEvents";
+import { logger } from "../../logger";
+
+export type BatchResult = {
+  result: unknown;
+  id: string;
+  type: string;
+};
+
+type TokenCountInput = {
+  model: Model;
+  text: unknown;
+};
+
+export type LegacyIngestionAccessScope = Omit<
+  ApiAccessScope,
+  "orgId" | "plan" | "rateLimitOverrides"
+>;
+
+type LegacyIngestionAuthHeaderVerificationResult =
+  | {
+      validKey: true;
+      scope: LegacyIngestionAccessScope;
+    }
+  | {
+      validKey: false;
+      error: string;
+    };
+
+export const handleBatch = async (
+  events: z.infer<typeof ingestionApiSchema>["batch"],
+  authCheck: LegacyIngestionAuthHeaderVerificationResult,
+  calculateTokenDelegate: (p: TokenCountInput) => number | undefined,
+) => {
+  logger.info(`handling ingestion ${events.length} events`);
+
+  if (!authCheck.validKey) throw new UnauthorizedError(authCheck.error);
+
+  const results: BatchResult[] = []; // Array to store the results
+
+  const errors: {
+    error: unknown;
+    id: string;
+    type: string;
+  }[] = []; // Array to store the errors
+
+  for (const singleEvent of events) {
+    try {
+      const result = await retry(async () => {
+        return await handleSingleEvent(
+          singleEvent,
+          authCheck.scope,
+          calculateTokenDelegate,
+        );
+      });
+      results.push({
+        result: result,
+        id: singleEvent.id,
+        type: singleEvent.type,
+      }); // Push each result into the array
+    } catch (error) {
+      // Handle or log the error if `handleSingleEvent` fails
+      logger.error("Error handling event:", error);
+      // Decide how to handle the error: rethrow, continue, or push an error object to results
+      // For example, push an error object:
+      errors.push({
+        error: error,
+        id: singleEvent.id,
+        type: singleEvent.type,
+      });
+    }
+  }
+
+  if (env.CLICKHOUSE_URL) {
+    try {
+      await enqueueIngestionEvents(authCheck.scope.projectId, events);
+      logger.info(`Added ${events.length} ingestion events to queue`);
+    } catch (err) {
+      logger.error("Error adding ingestion events to queue", err);
+    }
+  }
+
+  return { results, errors };
+};
+
+async function retry<T>(request: () => Promise<T>): Promise<T> {
+  return await backOff(request, {
+    numOfAttempts: env.LANGFUSE_ASYNC_INGESTION_PROCESSING === "true" ? 5 : 3,
+    retry: (e: Error, attemptNumber: number) => {
+      if (e instanceof UnauthorizedError || e instanceof ForbiddenError) {
+        logger.info("not retrying auth error");
+        return false;
+      }
+      logger.info(`retrying processing events ${attemptNumber}`);
+      return true;
+    },
+  });
+}
+
+const handleSingleEvent = async (
+  event: z.infer<typeof ingestionEvent>,
+  apiScope: LegacyIngestionAccessScope,
+  calculateTokenDelegate: (p: {
+    model: Model;
+    text: unknown;
+  }) => number | undefined,
+) => {
+  const { body } = event;
+  let restEvent = body;
+  if ("input" in body) {
+    // eslint-disable-next-line @typescript-eslint/no-unused-vars
+    const { input, ...rest } = body;
+    restEvent = rest;
+  }
+  if ("output" in restEvent) {
+    // eslint-disable-next-line @typescript-eslint/no-unused-vars
+    const { output, ...rest } = restEvent;
+    restEvent = rest;
+  }
+
+  logger.info(
+    `handling single event ${event.id} of type ${event.type}:  ${JSON.stringify({ body: restEvent })}`,
+  );
+
+  const cleanedEvent = ingestionEvent.parse(cleanEvent(event));
+
+  const { type } = cleanedEvent;
+
+  let processor: EventProcessor;
+  switch (type) {
+    case eventTypes.TRACE_CREATE:
+      processor = new TraceProcessor(cleanedEvent);
+      break;
+    case eventTypes.OBSERVATION_CREATE:
+    case eventTypes.OBSERVATION_UPDATE:
+    case eventTypes.EVENT_CREATE:
+    case eventTypes.SPAN_CREATE:
+    case eventTypes.SPAN_UPDATE:
+    case eventTypes.GENERATION_CREATE:
+    case eventTypes.GENERATION_UPDATE:
+      processor = new ObservationProcessor(
+        cleanedEvent,
+        calculateTokenDelegate,
+      );
+      break;
+    case eventTypes.SCORE_CREATE: {
+      processor = new ScoreProcessor(cleanedEvent);
+      break;
+    }
+    case eventTypes.SDK_LOG:
+      processor = new SdkLogProcessor(cleanedEvent);
+  }
+
+  // Deny access to non-score events if the access level is not "all"
+  // This is an additional safeguard to auth checks in EventProcessor
+  if (apiScope.accessLevel !== "all" && type !== eventTypes.SCORE_CREATE) {
+    throw new ForbiddenError("Access denied. Event type not allowed.");
+  }
+
+  return await processor.process(apiScope);
+};
+
+// cleans NULL characters from the event
+export function cleanEvent(obj: unknown): unknown {
+  if (typeof obj === "string") {
+    return obj.replace(/\u0000/g, "");
+  } else if (typeof obj === "object" && obj !== null) {
+    if (Array.isArray(obj)) {
+      return obj.map(cleanEvent);
+    } else {
+      // Here we assert that obj is a Record<string, unknown>
+      const objAsRecord = obj as Record<string, unknown>;
+      const newObj: Record<string, unknown> = {};
+      for (const key in objAsRecord) {
+        newObj[key] = cleanEvent(objAsRecord[key]);
+      }
+      return newObj;
+    }
+  } else {
+    return obj;
+  }
+}
+
+export const isNotNullOrUndefined = <T>(
+  val?: T | null,
+): val is Exclude<T, null | undefined> => !isUndefinedOrNull(val);
+
+export const isUndefinedOrNull = <T>(val?: T | null): val is undefined | null =>
+  val === undefined || val === null;
+
+export const sendToWorkerIfEnvironmentConfigured = async (
+  batchResults: BatchResult[],
+  projectId: string,
+): Promise<void> => {
+  const traceEvents: TraceUpsertEventType[] = batchResults
+    .filter((result) => result.type === eventTypes.TRACE_CREATE) // we only have create, no update.
+    .map((result) =>
+      result.result &&
+      typeof result.result === "object" &&
+      "id" in result.result
+        ? // ingestion API only gets traces for one projectId
+          { traceId: result.result.id as string, projectId }
+        : null,
+    )
+    .filter(isNotNullOrUndefined);
+
+  try {
+    if (env.NEXT_PUBLIC_LANGFUSE_CLOUD_REGION && redis) {
+      logger.info(`Sending ${traceEvents.length} events to worker via Redis`);
+
+      const queue = getTraceUpsertQueue();
+      if (!queue) {
+        logger.error("TraceUpsertQueue not initialized");
+        return;
+      }
+
+      await queue.addBulk(convertTraceUpsertEventsToRedisEvents(traceEvents));
+    } else if (
+      env.LANGFUSE_WORKER_HOST &&
+      env.LANGFUSE_WORKER_PASSWORD &&
+      env.NEXT_PUBLIC_LANGFUSE_CLOUD_REGION
+    ) {
+      logger.info(`Sending ${traceEvents.length} events to worker via HTTP`);
+      const body: EventBodyType = {
+        name: EventName.TraceUpsert,
+        payload: traceEvents,
+      };
+
+      if (traceEvents.length > 0) {
+        await fetch(`${env.LANGFUSE_WORKER_HOST}/api/events`, {
+          method: "POST",
+          headers: {
+            "Content-Type": "application/json",
+            Authorization:
+              "Basic " +
+              Buffer.from(
+                "admin" + ":" + env.LANGFUSE_WORKER_PASSWORD,
+              ).toString("base64"),
+          },
+          body: JSON.stringify(body),
+          signal: AbortSignal.timeout(8 * 1000),
+        });
+      }
+    }
+  } catch (error) {
+    logger.error("Error sending events to worker", error);
+  }
+};

packages/shared/src/server/ingestion/model-match.ts

@@ -0,0 +1,63 @@
+import { Model, Observation, Prisma } from "@prisma/client";
+import { prisma } from "../../db";
+
+export async function findModel(p: {
+  event: {
+    projectId: string;
+    model?: string;
+    unit?: string;
+    startTime?: Date;
+  };
+  existingDbObservation?: Observation;
+}): Promise<Model | null> {
+  const { event, existingDbObservation } = p;
+  // either get the model from the existing observation
+  // or match pattern on the user provided model name
+  const modelCondition = event.model
+    ? Prisma.sql`AND ${event.model} ~ match_pattern`
+    : existingDbObservation?.internalModel
+      ? Prisma.sql`AND model_name = ${existingDbObservation.internalModel}`
+      : undefined;
+  if (!modelCondition) return null;
+
+  // unit based on the current event or the existing observation, both can be undefined
+  const mergedUnit = event.unit ?? existingDbObservation?.unit;
+
+  const unitCondition = mergedUnit
+    ? Prisma.sql`AND unit = ${mergedUnit}`
+    : Prisma.empty;
+
+  const sql = Prisma.sql`
+    SELECT
+      id,
+      created_at AS "createdAt",
+      updated_at AS "updatedAt",
+      project_id AS "projectId",
+      model_name AS "modelName",
+      match_pattern AS "matchPattern",
+      start_date AS "startDate",
+      input_price AS "inputPrice",
+      output_price AS "outputPrice",
+      total_price AS "totalPrice",
+      unit, 
+      tokenizer_id AS "tokenizerId",
+      tokenizer_config AS "tokenizerConfig"
+    FROM
+      models
+    WHERE (project_id = ${event.projectId}
+      OR project_id IS NULL)
+    ${modelCondition}
+    ${unitCondition}
+    AND (start_date IS NULL OR start_date <= ${
+      event.startTime ? new Date(event.startTime) : new Date()
+    }::timestamp with time zone at time zone 'UTC')
+    ORDER BY
+      project_id ASC,
+      start_date DESC NULLS LAST
+    LIMIT 1
+  `;
+
+  const foundModels = await prisma.$queryRaw<Array<Model>>(sql);
+
+  return foundModels[0] ?? null;
+}

packages/shared/src/server/ingestion/types.ts

@@ -168,63 +168,39 @@ const BaseScoreBody = z.object({
 /**
  * ScoreBody exactly mirrors `PostScoresBody` in the public API. Please refer there for source of truth.
  */
-export const ScoreBody = z
-  .discriminatedUnion("dataType", [
-    BaseScoreBody.merge(
-      z.object({
-        value: z.number(),
-        dataType: z.literal("NUMERIC"),
-        configId: z.string().nullish(),
-      })
-    ),
-    BaseScoreBody.merge(
-      z.object({
-        value: z.string(),
-        dataType: z.literal("CATEGORICAL"),
-        configId: z.string().nullish(),
-      })
-    ),
-    BaseScoreBody.merge(
-      z.object({
-        value: z.number().refine((val) => val === 0 || val === 1, {
-          message: "Value must be either 0 or 1",
-        }),
-        dataType: z.literal("BOOLEAN"),
-        configId: z.string().nullish(),
-      })
-    ),
-    BaseScoreBody.merge(
-      z.object({
-        value: z.union([z.string(), z.number()]),
-        dataType: z.undefined(),
-        configId: z.string().nullish(),
-      })
-    ),
-  ])
-  .superRefine((data, ctx) => {
-    if (data.dataType) {
-      if (typeof data.value === "number") {
-        if (data.dataType === "CATEGORICAL") {
-          ctx.addIssue({
-            code: z.ZodIssueCode.custom,
-            message: `Value must be a string for data type ${data.dataType}`,
-          });
-        }
-      } else if (typeof data.value === "string") {
-        if (data.dataType === "NUMERIC") {
-          ctx.addIssue({
-            code: z.ZodIssueCode.custom,
-            message: `Value must be a number for data type ${data.dataType}`,
-          });
-        } else if (data.dataType === "BOOLEAN") {
-          ctx.addIssue({
-            code: z.ZodIssueCode.custom,
-            message: `Value must number equal to either 0 or 1 for data type ${data.dataType}`,
-          });
-        }
-      }
-    }
-  });
+export const ScoreBody = z.discriminatedUnion("dataType", [
+  BaseScoreBody.merge(
+    z.object({
+      value: z.number(),
+      dataType: z.literal("NUMERIC"),
+      configId: z.string().nullish(),
+    })
+  ),
+  BaseScoreBody.merge(
+    z.object({
+      value: z.string(),
+      dataType: z.literal("CATEGORICAL"),
+      configId: z.string().nullish(),
+    })
+  ),
+  BaseScoreBody.merge(
+    z.object({
+      value: z.number().refine((value) => value === 0 || value === 1, {
+        message:
+          "Value must be a number equal to either 0 or 1 for data type BOOLEAN",
+      }),
+      dataType: z.literal("BOOLEAN"),
+      configId: z.string().nullish(),
+    })
+  ),
+  BaseScoreBody.merge(
+    z.object({
+      value: z.union([z.string(), z.number()]),
+      dataType: z.undefined(),
+      configId: z.string().nullish(),
+    })
+  ),
+]);
 
 // LEGACY, only required for backwards compatibility
 export const LegacySpanPostSchema = z.object({
@@ -334,6 +310,16 @@ export const SdkLogEvent = z.object({
 
 // definitions for the ingestion API
 
+export const observationTypes = [
+  "observation-create",
+  "observation-update",
+  "generation-create",
+  "generation-update",
+  "span-create",
+  "span-update",
+  "event-create",
+];
+
 export const eventTypes = {
   TRACE_CREATE: "trace-create",
   SCORE_CREATE: "score-create",
@@ -343,7 +329,6 @@ export const eventTypes = {
   GENERATION_CREATE: "generation-create",
   GENERATION_UPDATE: "generation-update",
   SDK_LOG: "sdk-log",
-
   // LEGACY, only required for backwards compatibility
   OBSERVATION_CREATE: "observation-create",
   OBSERVATION_UPDATE: "observation-update",
@@ -358,6 +343,7 @@ export const traceEvent = base.extend({
   type: z.literal(eventTypes.TRACE_CREATE),
   body: TraceBody,
 });
+export type TraceEventType = z.infer<typeof traceEvent>;
 
 export const eventCreateEvent = base.extend({
   type: z.literal(eventTypes.EVENT_CREATE),
@@ -383,6 +369,7 @@ export const scoreEvent = base.extend({
   type: z.literal(eventTypes.SCORE_CREATE),
   body: ScoreBody,
 });
+export type ScoreEventType = z.infer<typeof scoreEvent>;
 export const sdkLogEvent = base.extend({
   type: z.literal(eventTypes.SDK_LOG),
   body: SdkLogEvent,
@@ -409,10 +396,35 @@ export const ingestionEvent = z.discriminatedUnion("type", [
   legacyObservationCreateEvent,
   legacyObservationUpdateEvent,
 ]);
+export type IngestionEventType = z.infer<typeof ingestionEvent>;
 
 export const ingestionBatchEvent = z.array(ingestionEvent);
+export type IngestionBatchEventType = z.infer<typeof ingestionBatchEvent>;
+
+export const ingestionEventWithProjectId = ingestionEvent.and(
+  z.object({ projectId: z.string() })
+);
+export type IngestionEventWithProjectIdType = z.infer<
+  typeof ingestionEventWithProjectId
+>;
 
 export const ingestionApiSchema = z.object({
   batch: ingestionBatchEvent,
   metadata: jsonSchema.nullish(),
 });
+
+export const ingestionApiSchemaWithProjectId = ingestionApiSchema.extend({
+  projectId: z.string(),
+});
+export type IngestionApiSchemaWithProjectId = z.infer<
+  typeof ingestionApiSchemaWithProjectId
+>;
+
+export type ObservationEvent =
+  | z.infer<typeof legacyObservationCreateEvent>
+  | z.infer<typeof legacyObservationUpdateEvent>
+  | z.infer<typeof eventCreateEvent>
+  | z.infer<typeof spanCreateEvent>
+  | z.infer<typeof spanUpdateEvent>
+  | z.infer<typeof generationCreateEvent>
+  | z.infer<typeof generationUpdateEvent>;

packages/shared/src/server/ingestion/validateAndInflateScore.ts

@@ -0,0 +1,160 @@
+import {
+  ScoreBodyWithoutConfig,
+  ScorePropsAgainstConfig,
+  validateDbScoreConfigSafe,
+  ValidatedScoreConfig,
+} from "../../../src";
+import { prisma, Score, ScoreDataType } from "../../db";
+
+import { InvalidRequestError, LangfuseNotFoundError } from "../../errors";
+
+type ValidateAndInflateScoreParams = {
+  projectId: string;
+  scoreId: string;
+  body: any;
+};
+
+export async function validateAndInflateScore(
+  params: ValidateAndInflateScoreParams,
+): Promise<Score> {
+  const { body, projectId } = params;
+
+  if (body.configId) {
+    const config = await prisma.scoreConfig.findFirst({
+      where: {
+        projectId,
+        id: body.configId,
+      },
+    });
+
+    if (!config || !validateDbScoreConfigSafe(config).success)
+      throw new LangfuseNotFoundError(
+        "The configId you provided does not match a valid config in this project",
+      );
+
+    validateConfigAgainstBody(body, config as ValidatedScoreConfig);
+
+    return inflateScoreBody({
+      ...params,
+      config: config as ValidatedScoreConfig,
+    });
+  }
+
+  const validation = ScoreBodyWithoutConfig.safeParse({
+    ...body,
+    dataType: body.dataType ?? inferDataType(body.value),
+  });
+
+  if (!validation.success) {
+    throw new InvalidRequestError(
+      `Ingested score value type not valid against provided data type. Provide numeric values for numeric and boolean scores, and string values for categorical scores.`,
+    );
+  }
+
+  return inflateScoreBody(params);
+}
+
+function inferDataType(value: string | number): ScoreDataType {
+  return typeof value === "number"
+    ? ScoreDataType.NUMERIC
+    : ScoreDataType.CATEGORICAL;
+}
+
+function mapStringValueToNumericValue(
+  config: ValidatedScoreConfig,
+  label: string,
+): number | null {
+  return (
+    config.categories?.find((category) => category.label === label)?.value ??
+    null
+  );
+}
+
+function inflateScoreBody(
+  params: ValidateAndInflateScoreParams & { config?: ValidatedScoreConfig },
+): Score {
+  const { body, projectId, scoreId, config } = params;
+
+  const relevantDataType = config?.dataType ?? body.dataType;
+  const scoreProps = { ...body, id: scoreId, projectId, source: "API" };
+
+  if (typeof body.value === "number") {
+    if (relevantDataType && relevantDataType === ScoreDataType.BOOLEAN) {
+      return {
+        ...scoreProps,
+        value: body.value,
+        stringValue: body.value === 1 ? "True" : "False",
+        dataType: ScoreDataType.BOOLEAN,
+      };
+    }
+
+    return {
+      ...scoreProps,
+      value: body.value,
+      dataType: ScoreDataType.NUMERIC,
+    };
+  }
+
+  return {
+    ...scoreProps,
+    value: config ? mapStringValueToNumericValue(config, body.value) : null,
+    stringValue: body.value,
+    dataType: ScoreDataType.CATEGORICAL,
+  };
+}
+
+function validateConfigAgainstBody(
+  body: any,
+  config: ValidatedScoreConfig,
+): void {
+  const { maxValue, minValue, categories, dataType: configDataType } = config;
+
+  if (body.dataType && body.dataType !== configDataType) {
+    throw new InvalidRequestError(
+      `Data type mismatch based on config: expected ${configDataType}, got ${body.dataType}`,
+    );
+  }
+
+  if (config.isArchived) {
+    throw new InvalidRequestError(
+      "Config is archived and cannot be used to create new scores. Please restore the config first.",
+    );
+  }
+
+  if (config.name !== body.name) {
+    throw new InvalidRequestError(
+      `Name mismatch based on config: expected ${config.name}, got ${body.name}`,
+    );
+  }
+
+  const relevantDataType = configDataType ?? body.dataType;
+
+  const dataTypeValidation = ScoreBodyWithoutConfig.safeParse({
+    ...body,
+    dataType: relevantDataType,
+  });
+
+  if (!dataTypeValidation.success) {
+    throw new InvalidRequestError(
+      `Ingested score body not valid against provided config data type.`,
+    );
+  }
+
+  const rangeValidation = ScorePropsAgainstConfig.safeParse({
+    value: body.value,
+    dataType: relevantDataType,
+    ...(maxValue !== null && maxValue !== undefined && { maxValue }),
+    ...(minValue !== null && minValue !== undefined && { minValue }),
+    ...(categories && { categories }),
+  });
+
+  if (!rangeValidation.success) {
+    const errorDetails = rangeValidation.error.errors
+      .map((error) => `${error.path.join(".")} - ${error.message}`)
+      .join(", ");
+
+    throw new InvalidRequestError(
+      `Ingested score body not valid against provided config: ${errorDetails}`,
+    );
+  }
+}

packages/shared/src/server/instrumentation/Readme.md

@@ -0,0 +1,24 @@
+# Instrumentation at Langfuse
+
+Throughout our applications we want to use as much Otel as possible. This helps us to be flexible choosing our observability backend, and we will benefit from features and packages built by the Otel community.
+
+## How to use Otel in your application
+
+- Use the `instrument` or `instrument` functions to wrap your functions with Otel instrumentation. This will automatically create spans for your functions and send them to the Otel collector. If an instrumented function throws, exceptions will be added to the span and the span will be marked as failed.
+- Use `recordGauge`, `recordCounter`, `recordHistogram` to record metrics. These will be sent to the Otel collector.
+
+## Configuration options
+
+- `web` and `worker` have an instrumentation.ts file, which configures otel for the application.
+- For trpc, we use `@baselime/trpc-opentelemetry-middleware` to enrich spans with trpc inputs and outputs.
+- When building adding new infrastructure, we should search for auto instumentations for our code base.
+
+## Library support
+
+- `dd-trace` has a direct dependency of `@opentelemetry/api` version < `1.9` [GH](https://github.com/DataDog/dd-trace-js/blob/ed9b0b30f7b0283579a9bf8c18e1f9deab18fecf/package.json#L81).
+- `@prisma/instrumentation` requires at least `@opentelemetry/api` version `1.8` [GH](https://github.com/prisma/prisma/blob/d780290b13754420abcfa5d7592f02049c6cc005/packages/instrumentation/package.json#L25)
+- We need to ensure that all other libraries we use are compatible with `@opentelemetry/api` version `1.8`. If we have a package using a later version, this will break the instrumentation and crash the container. We can check the versions with the following command:
+
+```bash
+pnpm --filter=web why @opentelemetry/api
+```

packages/shared/src/server/instrumentation/index.ts

@@ -0,0 +1,160 @@
+import * as opentelemetry from "@opentelemetry/api";
+import * as dd from "dd-trace";
+
+// type CallbackFn<T> = () => T;
+
+export type SpanCtx = {
+  name: string;
+  spanKind?: opentelemetry.SpanKind; // https://github.com/open-telemetry/opentelemetry-specification/blob/main/specification/trace/api.md#spankind
+  rootSpan?: boolean; // https://github.com/open-telemetry/opentelemetry-specification/blob/main/specification/overview.md#traces
+  traceScope?: string;
+};
+
+type AsyncCallbackFn<T> = () => Promise<T>;
+
+export async function instrumentAsync<T>(
+  ctx: SpanCtx,
+  callback: AsyncCallbackFn<T>
+): Promise<T> {
+  return await getTracer(ctx.traceScope ?? callback.name).startActiveSpan(
+    ctx.name,
+    {
+      root: ctx.rootSpan,
+      kind: ctx.spanKind,
+    },
+    async (span) => {
+      try {
+        const result = await callback();
+        span.end();
+        return result;
+      } catch (ex) {
+        traceException(ex as opentelemetry.Exception, span);
+        span.end();
+        throw ex;
+      }
+    }
+  );
+}
+
+type SyncCallbackFn<T> = () => T;
+
+export function instrumentSync<T>(
+  ctx: SpanCtx,
+  callback: SyncCallbackFn<T>
+): T {
+  return getTracer(ctx.traceScope ?? callback.name).startActiveSpan(
+    ctx.name,
+    {
+      root: ctx.rootSpan,
+      kind: ctx.spanKind,
+    },
+    (span) => {
+      try {
+        const result = callback();
+        span.end();
+        return result;
+      } catch (ex) {
+        traceException(ex as opentelemetry.Exception, span);
+        span.end();
+        throw ex;
+      }
+    }
+  );
+}
+
+export const getCurrentSpan = () => opentelemetry.trace.getActiveSpan();
+
+export const traceException = (
+  ex: unknown,
+  span?: opentelemetry.Span,
+  code?: string
+) => {
+  const activeSpan = span ?? getCurrentSpan();
+
+  if (!activeSpan) {
+    return;
+  }
+
+  const exception = {
+    code: code,
+    message:
+      ex instanceof Error
+        ? ex.message
+        : typeof ex === "object" && ex !== null && "message" in ex
+          ? JSON.stringify(ex.message)
+          : JSON.stringify(ex),
+    name:
+      ex instanceof Error
+        ? ex.name
+        : typeof ex === "object" && ex !== null && "name" in ex
+          ? JSON.stringify(ex.name)
+          : "Error",
+    stack:
+      ex instanceof Error
+        ? JSON.stringify(ex.stack)
+        : typeof ex === "object" && ex !== null && "stack" in ex
+          ? JSON.stringify(ex.stack)
+          : undefined,
+  };
+
+  // adds an otel event
+  activeSpan.recordException(exception);
+
+  //adds tags for datadog error tracking
+  activeSpan.setAttributes({
+    "error.stack": exception.stack,
+    "error.message": exception.message,
+    "error.type": exception.name,
+  });
+
+  activeSpan.setStatus({
+    code: opentelemetry.SpanStatusCode.ERROR,
+    message: exception.message,
+  });
+};
+
+export const addUserToSpan = (
+  attributes: { userId?: string; projectId?: string; email?: string },
+  span?: opentelemetry.Span
+) => {
+  const activeSpan = span ?? getCurrentSpan();
+
+  if (!activeSpan) {
+    return;
+  }
+
+  attributes.userId && activeSpan.setAttribute("user.id", attributes.userId);
+  attributes.email && activeSpan.setAttribute("user.email", attributes.email);
+  attributes.projectId &&
+    activeSpan.setAttribute("project.id", attributes.projectId);
+};
+
+export const getTracer = (name: string) => opentelemetry.trace.getTracer(name);
+
+export const recordGauge = (
+  stat: string,
+  value?: number | undefined,
+  tags?:
+    | {
+        [tag: string]: string | number;
+      }
+    | undefined
+) => {
+  dd.dogstatsd.gauge(stat, value, tags);
+};
+
+export const recordIncrement = (
+  stat: string,
+  value?: number | undefined,
+  tags?: { [tag: string]: string | number } | undefined
+) => {
+  dd.dogstatsd.increment(stat, value, tags);
+};
+
+export const recordHistogram = (
+  stat: string,
+  value?: number | undefined,
+  tags?: { [tag: string]: string | number } | undefined
+) => {
+  dd.dogstatsd.histogram(stat, value, tags);
+};

packages/shared/src/server/llm/types.ts

@@ -45,9 +45,13 @@ export const ZodModelConfig = z.object({
   top_p: z.coerce.number().optional(),
 });
 
+// NOTE: Update docs page when changing this!
 export const openAIModels = [
   "gpt-4o",
+  "gpt-4o-2024-08-06",
   "gpt-4o-2024-05-13",
+  "gpt-4o-mini",
+  "gpt-4o-mini-2024-07-18",
   "gpt-4-turbo-preview",
   "gpt-4-1106-preview",
   "gpt-4-0613",
@@ -64,6 +68,7 @@ export const openAIModels = [
 
 export type OpenAIModel = (typeof openAIModels)[number];
 
+// NOTE: Update docs page when changing this!
 export const anthropicModels = [
   "claude-3-5-sonnet-20240620",
   "claude-3-opus-20240229",

packages/shared/src/server/logger.ts

@@ -0,0 +1,33 @@
+import { env } from "../env";
+import winston from "winston";
+
+const getWinstonLogger = (
+  nodeEnv: "development" | "production" | "test",
+  minLevel = "info",
+) => {
+  const textLoggerFormat = winston.format.combine(
+    winston.format.errors({ stack: true }),
+    winston.format.timestamp(),
+    winston.format.align(),
+    winston.format.printf((info) => {
+      const logMessage = `${info.timestamp} ${info.level} ${info.message}`;
+      return info.stack ? `${logMessage}\n${info.stack}` : logMessage;
+    }),
+  );
+
+  const jsonLoggerFormat = winston.format.combine(
+    winston.format.errors({ stack: true }),
+    winston.format.timestamp(),
+    winston.format.json(),
+  );
+
+  const format =
+    env.LANGFUSE_LOG_FORMAT === "text" ? textLoggerFormat : jsonLoggerFormat;
+  return winston.createLogger({
+    level: minLevel,
+    format: format,
+    transports: [new winston.transports.Console()],
+  });
+};
+
+export const logger = getWinstonLogger(env.NODE_ENV, env.LANGFUSE_LOG_LEVEL);

packages/shared/src/server/orderByToPrisma.ts

@@ -2,8 +2,9 @@ import { z } from "zod";
 
 import { Prisma } from "@prisma/client";
 
-import type { ColumnDefinition } from "./tableDefinitions/types";
-import type { OrderByState } from "./interfaces/orderBy";
+import type { ColumnDefinition } from "../tableDefinitions/types";
+import type { OrderByState } from "../interfaces/orderBy";
+import { logger } from "./logger";
 
 /**
  * Convert orderBy to SQL ORDER BY clause
@@ -13,20 +14,20 @@ import type { OrderByState } from "./interfaces/orderBy";
  */
 export function orderByToPrismaSql(
   orderBy: OrderByState,
-  tableColumns: ColumnDefinition[]
+  tableColumns: ColumnDefinition[],
 ): Prisma.Sql {
   if (!orderBy) {
-    return Prisma.sql`ORDER BY t.timestamp DESC NULLS LAST`;
+    return Prisma.sql`ORDER BY t.timestamp DESC`;
   }
   // Get column definition to map column to internal name, e.g. "t.id"
   const col = tableColumns.find(
     // TODO: Only use id instead of name.
     // It's less error-prone & decouples data fetching from the human-readable UI labels
-    (c) => c.name === orderBy.column || c.id === orderBy.column
+    (c) => c.name === orderBy.column || c.id === orderBy.column,
   );
 
   if (!col) {
-    console.log("Invalid filter column", orderBy.column);
+    logger.warn("Invalid filter column", orderBy.column);
     throw new Error("Invalid filter column: " + orderBy.column);
   }
 
@@ -34,12 +35,12 @@ export function orderByToPrismaSql(
   const orderByOrder = z.enum(["ASC", "DESC"]);
   const order = orderByOrder.safeParse(orderBy.order);
   if (!order.success) {
-    console.log("Invalid order", orderBy.order);
+    logger.warn("Invalid order", orderBy.order);
     throw new Error("Invalid order: " + orderBy.order);
   }
 
   // Both column and order are safe, can use raw SQL
   return Prisma.raw(
-    `ORDER BY ${col.internal} ${order.data} ${orderBy.order === "DESC" ? "NULLS LAST" : "NULLS FIRST"}`
+    `ORDER BY ${col.internal} ${order.data} ${col.nullable ? (orderBy.order === "DESC" ? "NULLS LAST" : "NULLS FIRST") : ""}`,
   );
 }

packages/shared/src/server/queries/createSessionsAllQuery.ts

@@ -2,10 +2,10 @@ import { z } from "zod";
 
 import { Prisma } from "@prisma/client";
 import { tableColumnsToSqlFilterAndPrefix } from "../filterToPrisma";
-import { singleFilter } from "../interfaces/filters";
-import { orderBy } from "../interfaces/orderBy";
+import { singleFilter } from "../../interfaces/filters";
+import { orderBy } from "../../interfaces/orderBy";
 import { orderByToPrismaSql } from "../orderByToPrisma";
-import { sessionsViewCols } from "../tableDefinitions/index";
+import { sessionsViewCols } from "../../tableDefinitions";
 
 const GetSessionTableSQLParamsSchema = z.object({
   projectId: z.string(),
@@ -16,8 +16,13 @@ const GetSessionTableSQLParamsSchema = z.object({
 });
 type GetSessionTableSQLParams = z.infer<typeof GetSessionTableSQLParamsSchema>;
 
-export const getSessionTableSQL = (
-  params: GetSessionTableSQLParams
+export const createSessionsAllQuery = (
+  select: Prisma.Sql,
+  params: GetSessionTableSQLParams,
+  options?: {
+    ignoreOrderBy?: boolean; // used by session.metrics and session.all.totalCount
+    sessionIdList?: string[]; // used by session.metrics
+  },
 ): Prisma.Sql => {
   const { projectId, filter, orderBy, page, limit } =
     GetSessionTableSQLParamsSchema.parse(params);
@@ -25,26 +30,17 @@ export const getSessionTableSQL = (
   const filterCondition = tableColumnsToSqlFilterAndPrefix(
     filter ?? [],
     sessionsViewCols,
-    "sessions"
+    "sessions",
   );
   const orderByCondition = orderByToPrismaSql(orderBy, sessionsViewCols);
 
+  const sessionIdFilter = options?.sessionIdList
+    ? Prisma.sql`AND s.id IN (${Prisma.join(options?.sessionIdList)})`
+    : Prisma.sql``;
+
   const sql = Prisma.sql`
       SELECT
-        s.id,
-        s. "created_at" AS "createdAt",
-        s.bookmarked,
-        s.public,
-        t. "userIds",
-        t. "countTraces",
-        o. "sessionDuration",
-        o. "totalCost" AS "totalCost",
-        o. "inputCost" AS "inputCost",
-        o. "outputCost" AS "outputCost",
-        o. "promptTokens" AS "promptTokens",
-        o. "completionTokens" AS "completionTokens",
-        o. "totalTokens" AS "totalTokens",
-        (count(*) OVER ())::int AS "totalCount"
+        ${select}
       FROM
         trace_sessions AS s
         LEFT JOIN LATERAL (
@@ -54,9 +50,12 @@ export const getSessionTableSQL = (
             MIN(t. "timestamp") AS "min_timestamp",
             array_agg(t.id) AS "traceIds",
             array_agg(DISTINCT t.user_id) AS "userIds",
-            count(t.id)::int AS "countTraces"
+            count(t.id)::int AS "countTraces",
+            array_agg(DISTINCT u.tag) AS "tags"
           FROM
             traces t
+          LEFT JOIN LATERAL (
+            SELECT DISTINCT UNNEST(t.tags) AS tag) AS u ON TRUE
           WHERE
             t.project_id = ${projectId}
             AND t.session_id = s.id
@@ -79,7 +78,8 @@ export const getSessionTableSQL = (
       WHERE
         s. "project_id" = ${projectId}
         ${filterCondition}
-      ${orderByCondition}
+        ${sessionIdFilter}
+      ${options?.ignoreOrderBy ? Prisma.sql`` : orderByCondition}
       LIMIT ${limit}
       OFFSET ${page * limit}
     `;

packages/shared/src/server/queries/index.ts

@@ -0,0 +1 @@
+export { createSessionsAllQuery } from "./createSessionsAllQuery";

packages/shared/src/server/queues.ts

@@ -1,11 +1,25 @@
 import { z } from "zod";
+import { ingestionBatchEvent } from ".";
 
 export enum EventName {
   TraceUpsert = "TraceUpsert",
   BatchExport = "BatchExport",
   EvaluationExecution = "EvaluationExecution",
+  LegacyIngestion = "LegacyIngestion",
+  CloudUsageMetering = "CloudUsageMetering",
 }
 
+export const LegacyIngestionEvent = z.object({
+  data: ingestionBatchEvent,
+  authCheck: z.object({
+    validKey: z.literal(true),
+    scope: z.object({
+      projectId: z.string(),
+      accessLevel: z.enum(["all", "scores"]),
+    }),
+  }),
+});
+
 export const BatchExportJobSchema = z.object({
   projectId: z.string(),
   batchExportId: z.string(),
@@ -22,6 +36,7 @@ export const EvalExecutionEvent = z.object({
 export type BatchExportJobType = z.infer<typeof BatchExportJobSchema>;
 export type TraceUpsertEventType = z.infer<typeof TraceUpsertEventSchema>;
 export type EvalExecutionEventType = z.infer<typeof EvalExecutionEvent>;
+export type LegacyIngestionEventType = z.infer<typeof LegacyIngestionEvent>;
 
 export const EventBodySchema = z.union([
   z.object({
@@ -44,6 +59,9 @@ export enum QueueName {
   EvaluationExecution = "evaluation-execution-queue", // Worker executes Evals
   BatchExport = "batch-export-queue",
   RepeatQueue = "repeat-queue",
+  IngestionFlushQueue = "ingestion-flush-queue",
+  LegacyIngestionQueue = "legacy-ingestion-queue",
+  CloudUsageMeteringQueue = "cloud-usage-metering-queue",
 }
 
 export enum QueueJobs {
@@ -51,6 +69,9 @@ export enum QueueJobs {
   EvaluationExecution = "evaluation-execution-job",
   BatchExportJob = "batch-export-job",
   EnqueueBatchExportJobs = "enqueue-batch-export-jobs",
+  FlushIngestionEntity = "flush-ingestion-entity",
+  LegacyIngestionJob = "legacy-ingestion-job",
+  CloudUsageMeteringJob = "cloud-usage-metering-job",
 }
 
 export type TQueueJobTypes = {
@@ -72,4 +93,10 @@ export type TQueueJobTypes = {
     payload: BatchExportJobType;
     name: QueueJobs.BatchExportJob;
   };
+  [QueueName.LegacyIngestionQueue]: {
+    timestamp: Date;
+    id: string;
+    payload: LegacyIngestionEventType;
+    name: QueueJobs.LegacyIngestionJob;
+  };
 };

packages/shared/src/server/redis/batch-export.ts

@@ -0,0 +1,25 @@
+import { Queue } from "bullmq";
+import { QueueName, TQueueJobTypes } from "../queues";
+import { createNewRedisInstance } from "./redis";
+
+let batchExportQueue: Queue<TQueueJobTypes[QueueName.BatchExport]> | null =
+  null;
+
+export const getBatchExportQueue = () => {
+  if (batchExportQueue) return batchExportQueue;
+
+  const connection = createNewRedisInstance();
+
+  batchExportQueue = connection
+    ? new Queue<TQueueJobTypes[QueueName.BatchExport]>(QueueName.BatchExport, {
+        connection: connection,
+        defaultJobOptions: {
+          removeOnComplete: true,
+          removeOnFail: 100,
+          attempts: 2,
+        },
+      })
+    : null;
+
+  return batchExportQueue;
+};

packages/shared/src/server/redis/ingestionFlushQueue.ts

@@ -0,0 +1,29 @@
+import { Queue } from "bullmq";
+
+import { env } from "../../env";
+import { createNewRedisInstance } from "../redis/redis";
+import { QueueName } from "../queues";
+
+export type IngestionFlushQueue = Queue<null>;
+
+let ingestionFlushQueue: IngestionFlushQueue | null = null;
+
+export const getIngestionFlushQueue = () => {
+  if (ingestionFlushQueue) return ingestionFlushQueue;
+
+  const connection = createNewRedisInstance();
+
+  ingestionFlushQueue = connection
+    ? new Queue<null>(QueueName.IngestionFlushQueue, {
+        connection: connection,
+        defaultJobOptions: {
+          removeOnComplete: true, // Important: If not true, new jobs for that ID would be ignored as jobs in the complete set are still considered as part of the queue
+          removeOnFail: 1_000,
+          delay: env.LANGFUSE_INGESTION_FLUSH_DELAY_MS,
+          attempts: env.LANGFUSE_INGESTION_FLUSH_ATTEMPTS,
+        },
+      })
+    : null;
+
+  return ingestionFlushQueue;
+};

packages/shared/src/server/redis/legacy-ingestion.ts

@@ -0,0 +1,33 @@
+import { Queue } from "bullmq";
+import { QueueName, TQueueJobTypes } from "../queues";
+import { createNewRedisInstance } from "./redis";
+
+export class LegacyIngestionQueue {
+  private static instance: Queue<
+    TQueueJobTypes[QueueName.LegacyIngestionQueue]
+  > | null = null;
+
+  public static getInstance(): Queue<
+    TQueueJobTypes[QueueName.LegacyIngestionQueue]
+  > | null {
+    if (LegacyIngestionQueue.instance) return LegacyIngestionQueue.instance;
+
+    const newRedis = createNewRedisInstance({ enableOfflineQueue: false });
+
+    LegacyIngestionQueue.instance = newRedis
+      ? new Queue<TQueueJobTypes[QueueName.LegacyIngestionQueue]>(
+          QueueName.LegacyIngestionQueue,
+          {
+            connection: newRedis,
+            defaultJobOptions: {
+              removeOnComplete: true,
+              removeOnFail: 100,
+              attempts: 5,
+            },
+          },
+        )
+      : null;
+
+    return LegacyIngestionQueue.instance;
+  }
+}

packages/shared/src/server/redis/redis.ts

@@ -0,0 +1,41 @@
+import Redis, { RedisOptions } from "ioredis";
+import { env } from "../../env";
+import { logger } from "../logger";
+
+export const createNewRedisInstance = (
+  additionalOptions: Partial<RedisOptions> = {},
+) => {
+  return env.REDIS_CONNECTION_STRING
+    ? new Redis(env.REDIS_CONNECTION_STRING, {
+        maxRetriesPerRequest: null,
+        enableAutoPipelining: env.REDIS_ENABLE_AUTO_PIPELINING === "true",
+        ...additionalOptions,
+      })
+    : env.REDIS_HOST
+      ? new Redis({
+          host: String(env.REDIS_HOST),
+          port: Number(env.REDIS_PORT),
+          password: String(env.REDIS_AUTH),
+          maxRetriesPerRequest: null, // Set to `null` to disable retrying
+          enableAutoPipelining: env.REDIS_ENABLE_AUTO_PIPELINING === "true",
+          ...additionalOptions,
+        })
+      : null;
+};
+
+const createRedisClient = () => {
+  try {
+    return createNewRedisInstance();
+  } catch (e) {
+    logger.error("Failed to connect to redis", e);
+    return null;
+  }
+};
+declare global {
+  // eslint-disable-next-line no-var
+  var redis: undefined | ReturnType<typeof createRedisClient>;
+}
+
+export const redis = globalThis.redis ?? createRedisClient();
+
+if (env.NODE_ENV !== "production") globalThis.redis = redis;

packages/shared/src/server/redis/trace-upsert.ts

@@ -0,0 +1,71 @@
+import { randomUUID } from "crypto";
+import {
+  QueueJobs,
+  QueueName,
+  TQueueJobTypes,
+  TraceUpsertEventType,
+} from "../queues";
+import { Queue } from "bullmq";
+import { createNewRedisInstance } from "./redis";
+
+let traceUpsertQueue: Queue<TQueueJobTypes[QueueName.TraceUpsert]> | null =
+  null;
+
+export const getTraceUpsertQueue = () => {
+  if (traceUpsertQueue) return traceUpsertQueue;
+
+  const connection = createNewRedisInstance();
+
+  traceUpsertQueue = connection
+    ? new Queue<TQueueJobTypes[QueueName.TraceUpsert]>(QueueName.TraceUpsert, {
+        connection: connection,
+        defaultJobOptions: {
+          removeOnComplete: 100, // Important: If not true, new jobs for that ID would be ignored as jobs in the complete set are still considered as part of the queue
+          removeOnFail: 1_000,
+        },
+      })
+    : null;
+
+  return traceUpsertQueue;
+};
+
+export function convertTraceUpsertEventsToRedisEvents(
+  events: TraceUpsertEventType[]
+) {
+  const uniqueTracesPerProject = events.reduce((acc, event) => {
+    if (!acc.get(event.projectId)) {
+      acc.set(event.projectId, new Set());
+    }
+    acc.get(event.projectId)?.add(event.traceId);
+    return acc;
+  }, new Map<string, Set<string>>());
+
+  const jobs = [...uniqueTracesPerProject.entries()]
+    .map((tracesPerProject) => {
+      const [projectId, traceIds] = tracesPerProject;
+
+      return [...traceIds].map((traceId) => ({
+        name: QueueJobs.TraceUpsert,
+        data: {
+          payload: {
+            projectId,
+            traceId,
+          },
+          id: randomUUID(),
+          timestamp: new Date(),
+          name: QueueJobs.TraceUpsert as const,
+        },
+        opts: {
+          removeOnFail: 1_000,
+          removeOnComplete: true,
+          attempts: 5,
+          backoff: {
+            type: "exponential",
+            delay: 1000,
+          },
+        },
+      }));
+    })
+    .flat();
+  return jobs;
+}

packages/shared/src/server/services/PromptService.ts

@@ -0,0 +1,246 @@
+import { Prompt, PrismaClient } from "@prisma/client";
+import { Redis } from "ioredis";
+import { env } from "../../env";
+import { logger } from "../logger";
+
+export class PromptService {
+  private cacheEnabled: boolean;
+  private ttlSeconds: number;
+
+  constructor(
+    private prisma: PrismaClient,
+    private redis: Redis | null,
+    private metricIncrementer?: // used for otel metrics
+    (name: string, value?: number) => void,
+    cacheEnabled?: boolean, // used for testing
+  ) {
+    this.cacheEnabled =
+      Boolean(redis) &&
+      (cacheEnabled || env.LANGFUSE_CACHE_PROMPT_ENABLED === "true");
+
+    this.ttlSeconds = env.LANGFUSE_CACHE_PROMPT_TTL_SECONDS;
+  }
+
+  public async getPrompt(params: PromptParams): Promise<Prompt | null> {
+    if (await this.shouldUseCache(params)) {
+      const cachedPrompt = await this.getCachedPrompt(params);
+
+      this.incrementMetric(
+        cachedPrompt ? Metrics.PromptCacheHit : Metrics.PromptCacheMiss,
+      );
+
+      if (cachedPrompt) {
+        this.logInfo("Returning cached prompt for params", params);
+
+        return cachedPrompt;
+      }
+    }
+
+    const dbPrompt = await this.getDbPrompt(params);
+
+    if ((await this.shouldUseCache(params)) && dbPrompt) {
+      await this.cachePrompt({ ...params, prompt: dbPrompt });
+
+      this.logInfo("Successfully cached prompt for params", params);
+    }
+
+    this.logInfo("Returning DB prompt for params", params);
+
+    return dbPrompt;
+  }
+
+  private async getDbPrompt(params: PromptParams): Promise<Prompt | null> {
+    const { projectId, promptName, version, label } = params;
+
+    if (version) {
+      return await this.prisma.prompt.findFirst({
+        where: {
+          projectId,
+          name: promptName,
+          version,
+        },
+      });
+    }
+
+    if (label) {
+      return await this.prisma.prompt.findFirst({
+        where: {
+          projectId,
+          name: promptName,
+          labels: {
+            has: label,
+          },
+        },
+      });
+    }
+
+    this.logError("Invalid prompt params", params);
+
+    return null;
+  }
+
+  private async shouldUseCache(params: PromptParams): Promise<boolean> {
+    if (!this.cacheEnabled) return false;
+
+    const isLocked = await this.isCacheLocked(params);
+
+    if (isLocked) {
+      this.logInfo("Cache is locked for params", params);
+    }
+
+    return !isLocked;
+  }
+
+  private async getCachedPrompt(params: PromptParams): Promise<Prompt | null> {
+    try {
+      const key = this.getCacheKey(params);
+      const value = await this.redis?.getex(key, "EX", this.ttlSeconds);
+
+      if (value) return JSON.parse(value) as Prompt;
+    } catch (e) {
+      this.logError("Error getting cached prompt", e);
+    }
+
+    return null;
+  }
+
+  private async cachePrompt(params: PromptParams & { prompt: Prompt }) {
+    try {
+      const keyIndexKey = this.getKeyIndexKey(params);
+      const key = this.getCacheKey(params);
+      const value = JSON.stringify(params.prompt);
+
+      await this.redis?.sadd(keyIndexKey, key);
+      await this.redis?.set(key, value, "EX", this.ttlSeconds);
+    } catch (e) {
+      this.logError("Error caching prompt", e);
+    }
+  }
+
+  public async lockCache(
+    params: Pick<PromptParams, "projectId" | "promptName">,
+  ): Promise<void> {
+    if (!this.cacheEnabled) return;
+
+    const lockKey = this.getLockKey(params);
+
+    try {
+      await this.redis?.setex(lockKey, 30, "locked");
+    } catch (e) {
+      this.logError("Error locking cache key prefix", lockKey, e);
+
+      throw e;
+    }
+  }
+
+  public async unlockCache(
+    params: Pick<PromptParams, "projectId" | "promptName">,
+  ): Promise<void> {
+    if (!this.cacheEnabled) return;
+
+    const lockKey = this.getLockKey(params);
+
+    try {
+      await this.redis?.del(lockKey);
+    } catch (e) {
+      this.logError("Error unlocking cache key prefix", lockKey, e);
+
+      // Don't re-throw error as lock TTL is short and it's not critical
+    }
+  }
+
+  private async isCacheLocked(
+    params: Pick<PromptParams, "projectId" | "promptName">,
+  ): Promise<boolean> {
+    const lockKey = this.getLockKey(params);
+
+    try {
+      return Boolean(await this.redis?.exists(lockKey));
+    } catch (e) {
+      this.logError("Error checking if cache is locked", lockKey, e);
+
+      return false;
+    }
+  }
+
+  private getLockKey(
+    params: Pick<PromptParams, "projectId" | "promptName">,
+  ): string {
+    // Important to *pre*fix LOCK as otherwise it would be deleted by deleteKeysByPrefix
+    return `LOCK:${this.getCacheKeyPrefix(params)}`;
+  }
+
+  public async invalidateCache(
+    params: Pick<PromptParams, "projectId" | "promptName">,
+  ): Promise<void> {
+    if (!this.cacheEnabled) return;
+
+    const cacheKeyPrefix = this.getCacheKeyPrefix(params);
+
+    try {
+      const startTime = Date.now();
+      this.logInfo("Invalidating cache for prefix", cacheKeyPrefix);
+
+      const keyIndexKey = this.getKeyIndexKey(params);
+      const keys = await this.redis?.smembers(keyIndexKey);
+
+      // Delete all keys for the prefix and the key index
+      await this.redis?.del([...(keys ?? []), keyIndexKey]);
+
+      this.logInfo(
+        `Cache invalidated for prefix ${cacheKeyPrefix} in ${Date.now() - startTime}ms`,
+      );
+    } catch (e) {
+      this.logError("Error deleting keys for prefix", cacheKeyPrefix, e);
+
+      throw e;
+    }
+  }
+
+  private getCacheKey(params: PromptParams): string {
+    const prefix = this.getCacheKeyPrefix(params);
+
+    return `${prefix}:${params.version ?? params.label}`;
+  }
+
+  private getCacheKeyPrefix(
+    params: Pick<PromptParams, "projectId" | "promptName">,
+  ): string {
+    return `prompt:${params.projectId}:${params.promptName}`;
+  }
+
+  private getKeyIndexKey(
+    params: Pick<PromptParams, "projectId" | "promptName">,
+  ): string {
+    return `prompt_key_index:${params.projectId}:${params.promptName}`;
+  }
+
+  private logError(message: string, ...args: any[]) {
+    logger.error(`[PromptService] ${message}`, ...args);
+  }
+
+  private logInfo(message: string, ...args: any[]) {
+    logger.info(`[PromptService] ${message}`, ...args);
+  }
+
+  private incrementMetric(name: Metrics, value: number = 1) {
+    try {
+      this.metricIncrementer?.(name, value);
+    } catch (e) {
+      this.logError("Error incrementing metric", name, e);
+    }
+  }
+}
+
+type PromptParams = {
+  projectId: string;
+  promptName: string;
+} & (
+  | { version: number; label: undefined }
+  | { version: null | undefined; label: string }
+);
+
+enum Metrics {
+  PromptCacheHit = "prompt_cache_hit",
+  PromptCacheMiss = "prompt_cache_miss",
+}

packages/shared/src/server/services/S3StorageService.ts

@@ -2,6 +2,7 @@ import type { Readable } from "stream";
 import { GetObjectCommand, S3Client } from "@aws-sdk/client-s3";
 import { Upload } from "@aws-sdk/lib-storage";
 import { getSignedUrl } from "@aws-sdk/s3-request-presigner";
+import { logger } from "../logger";
 
 type UploadFile = {
   fileName: string;
@@ -56,15 +57,14 @@ export class S3StorageService {
 
       return { signedUrl };
     } catch (err) {
-      console.error(err);
-
+      logger.error(err);
       throw new Error("Failed to upload to S3 or generate signed URL");
     }
   }
 
   private async getSignedUrl(
     fileName: string,
-    ttlSeconds: number
+    ttlSeconds: number,
   ): Promise<string> {
     try {
       return await getSignedUrl(
@@ -74,7 +74,7 @@ export class S3StorageService {
           Key: fileName,
           ResponseContentDisposition: `attachment; filename="${fileName}"`,
         }),
-        { expiresIn: ttlSeconds }
+        { expiresIn: ttlSeconds },
       );
     } catch (err) {
       throw Error("Failed to generate signed URL");

packages/shared/src/server/services/email/batchExportSuccess/sendBatchExportSuccessEmail.ts

@@ -3,6 +3,7 @@ import { parseConnectionUrl } from "nodemailer/lib/shared/index.js";
 import { render } from "@react-email/render";
 
 import { BatchExportSuccessEmailTemplate } from "./BatchExportSuccessEmailTemplate";
+import { logger } from "../../../logger";
 
 type SendBatchExportSuccessParams = {
   env: Partial<
@@ -24,8 +25,7 @@ export const sendBatchExportSuccessEmail = async ({
   expiresInHours,
 }: SendBatchExportSuccessParams) => {
   if (!env.EMAIL_FROM_ADDRESS || !env.SMTP_CONNECTION_URL) {
-    console.error("Missing environment variables for sending email.");
-
+    logger.error("Missing environment variables for sending email.");
     return;
   }
 
@@ -38,7 +38,7 @@ export const sendBatchExportSuccessEmail = async ({
         userName,
         batchExportName,
         expiresInHours,
-      })
+      }),
     );
 
     await mailer.sendMail({
@@ -51,6 +51,6 @@ export const sendBatchExportSuccessEmail = async ({
       html: htmlTemplate,
     });
   } catch (error) {
-    console.error(error);
+    logger.error(error);
   }
 };