mirror of https://github.com/hanzoai/insights-js.git synced 2026-11-03 00:53:53 +00:00

Files

T

Richard SolomouandGitHub 54108e8385 docs: add AI provider examples for all LLM analytics integrations (#3326 )

* docs: add AI provider examples for all LLM analytics integrations

Add 20 new example directories covering every provider documented in
our LLM analytics onboarding docs. This includes OpenAI-compatible
providers (Groq, DeepSeek, Mistral, xAI, Together AI, Ollama, Cohere,
Hugging Face, Perplexity, Cerebras, Fireworks AI, OpenRouter, Helicone,
Vercel AI Gateway, Azure OpenAI, Portkey) and custom integrations
(Instructor, LangGraph, AWS Bedrock, Convex).

* fix: use max_tokens instead of max_completion_tokens for Mistral example

Mistral's API doesn't support max_completion_tokens, causing 422 errors.

* fix: import AzureOpenAI from @posthog/ai, not @posthog/ai/openai

* fix: use Resource constructor instead of resourceFromAttributes for OTel v1

* style: format all AI examples with prettier

2026-04-07 11:07:35 +03:00

385 B

Raw Permalink Blame History

AWS Bedrock + PostHog AI Examples

Track AWS Bedrock LLM calls with PostHog via OpenTelemetry instrumentation.

Setup

pnpm install
cp .env.example .env
# Fill in your API keys in .env
# Ensure your AWS credentials are configured

Examples

chat.ts - Bedrock Converse API with OpenTelemetry tracing to PostHog

Run

source .env
npx tsx chat.ts

385 B Raw Permalink Blame History

AWS Bedrock + PostHog AI Examples

Setup

Examples

Run

385 B

Raw Permalink Blame History