What is Maia?

Maia is a self-hosted DAG workflow orchestration and execution service for long-running automation—observable, debuggable, and replayable.

• Persistence: state and outputs are persisted with SQLite (retained and traceable)
• Observability: real-time logs/event streams (SSE) with replay
• Isolated execution: optional Runner + Sandbox container isolation (recommended for production)
• Composable: each step has explicit inputs/outputs and can produce artifacts
• Optional agent: helps generate/refine workflows (opt-in)

Control plane vs execution plane

In the recommended Docker deployment, Maia is split into two services:

• App (control plane): UI + scheduler/state machine + SQLite + real-time SSE
• Runner (execution plane): starts/stops sandbox containers and streams stdout/stderr back to App

This architecture lets you scale and operate “orchestration/state” separately from “execution/isolation”.

When do you need Maia?

• Your jobs run for a long time (minutes/hours/days) and need reliable state persistence
• You want full observability for every run (real-time logs, outputs, inputs, artifacts)
• After failures, you want to retry/rerun/replay instead of relying on “black-box scripts”
• You want to break automation into composable steps, with visualization and debugging in the UI