Latitude

An open-source AI agent monitoring platform you can self-host with Docker or Kubernetes to find failures, dispatch coding agents and check fixes against traces.

Latitude is an AI agent monitoring platform for teams investigating production failures. It links recurring problems to the conversations and tool calls behind them, then gives coding agents the context to work on a fix.

Session tracing: Capture multi-turn conversations, tool calls, errors, costs and execution paths, then find affected sessions through semantic search, exact text and metadata filters.
Failure detection: Group similar failures into tracked issues with examples and trends, and define custom signals for specific behaviors, policies or quality checks.
Alerts and fixes: Send issue alerts through Slack, email or webhooks, and dispatch Claude Code or Cursor with sample traces and context to implement a fix and open a pull request.
Regression checks: Build versioned evaluation datasets from validated production traces, add human annotations, and replay fixes against failing traces.

Latitude accepts OpenTelemetry traces and integrates with OpenAI, Anthropic, Bedrock, Vercel AI SDK and LangChain. Its MCP server and CLI let coding agents access workspace and project controls.

The software is open source under MIT and can run on your own server with Docker Compose or on Kubernetes. A hosted service is also available, with hosted data stored and processed in Europe. Self-hosting controls where the monitoring platform runs; your application's model provider is a separate choice.