Rhesis Documentation
Structured feedback and evals for AI agents.
Rhesis is the collaboration layer between the people building an AI agent and the people who know what it should answer. Connect your agent, share the link with your team, and turn their feedback into tests and metrics that run on every change.
Getting Started
Setup Environment
Run Rhesis as a managed service, locally with Docker, or self-hosted. Setup Environment Guide →
Connect Application
Use the SDK connector and your process opens an outbound WebSocket, so the agent stays on your laptop or inside your VPC with no public URL. If it already serves a public REST endpoint, register that URL instead. Connect Application Guide →
Run Evaluations
Share the link with domain experts and product managers. They chat with the live agent, turn conversations into tests, and leave pass/fail verdicts and comments down to the individual metric or conversation turn — then you read it all back from the SDK, the API, or MCP. Run Evaluations Guide →