Monitor your OpenAI agents with Datadog LLM Observability

This Datadog article explains how annotating LLM traces—adding metadata about inputs, outputs, and context—is crucial for understanding and improving LLM performance. By enriching traces with this information, teams can pinpoint the root cause of issues like hallucination or toxicity, and ultimately refine prompts, models, and retrieval-augmented generation (RAG) pipelines for better quality outputs. Datadog's LLM Observability features facilitate this annotation process and provide tools to analyze the resulting data.