Skip to main content
Maitai is a platform for fine-tuning, testing, and monitoring LLM-powered applications. You can route your LLM requests through Maitai (recommended), or keep your existing client and send Maitai the request/response for indexing. Maitai organizes traffic by Application and Intent, runs Sentinels to detect faults, and gives you Portal workflows to iterate safely with Test Sets and Test Runs.

​
Start here

  • If you want the fastest setup: Quickstart
  • If you already use an OpenAI-compatible SDK and want a drop-in integration: Base URL integration
  • If you want a mental model of the resources you’ll see in the Portal: Core Concepts

​
What you’ll get out of Maitai

  • Observe: sessions/requests, latency + cost, and evaluation outcomes in the Portal
  • Test: regression suites (Test Sets) and benchmarks (Test Runs)
  • Build: fine-tuning datasets/runs, and Agents for tool-using workflows