Getting Started with Noodler
Noodler is an open-source AI observability platform built on OpenTelemetry. It helps teams understand how their AI applications perform by capturing real user interactions and model behavior.
Overview
Noodler is your home for understanding how your AI application performs.
That includes both reactive workflows, like monitoring and debugging, and active workflows, like improving prompts, models, and behavior over time. To support this, Noodler is built from a small set of focused building blocks.
With multiple components, it’s fair to wonder where to start. In many systems, the answer would be “it depends.” With Noodler, it doesn’t.
Today, Noodler focuses on a single capability: observability.
Observability lets you see how your AI behaves in the real world. By capturing real interactions — what users ask and how the AI responds — you gain a clear picture of both user intent and system behavior.
You can’t improve what you can’t see.
Next Steps
- Self-Host Noolder.
- Install the SDK.
- Get to understand Traces — the building block of observability.