Phoenix

Provides total visibility into LLM application data with automatic or manual instrumentation.

Uncovers semantically similar questions, document chunks, and responses using embeddings to isolate poor performance.

Enables evaluation, experimentation, and optimization of AI products in real time.

Visualizes complex LLM decision-making and flags when and where models fail or produce poor responses.

Offers a fast, flexible sandbox for prompt and model iteration, output visualization, and debugging.

Leverages a high-speed evaluation library with pre-built, customizable templates for any task.

Incorporates human feedback into the evaluation process to enhance model assessment.

Uses OpenTelemetry for seamless setup, full transparency, and freedom from vendor lock-in.

Provided as a fully open-source solution, allowing for self-hosting without feature restrictions.

Works with all major LLM tools and frameworks, supporting easy integration into existing workflows.