Introduction

Maestro is an open-source multi-agent orchestration platform. It handles the infrastructure for running AI agents—routing, memory, tools, and observability—so you can focus on what your agents do.

Why Maestro?

Building production AI agents involves more than just prompting. You need:

• Routing to direct requests to the right specialized agent
• Memory to persist conversations across sessions
• Tools to give agents real capabilities
• Channels to deploy agents to multiple interfaces
• Observability to track costs and debug issues

Maestro handles all of this out of the box.

Key Features

Multi-Agent Routing

The orchestrator analyzes requests and routes them to specialized agents (coder, assistant, or your custom agents).

Persistent Memory

SQLite-backed conversations survive restarts. No configuration required.

Channel Abstraction

Write agent logic once, deploy to Telegram, CLI, REST API, and more.

Dynamic Agents

Create new agents at runtime through conversation. No restart required.

How It Works

User Message → Channel (Telegram/CLI/API)
     ↓
Orchestrator (analyzes intent)
     ↓
Specialized Agent (personal-assistant, coder, custom)
     ↓
Response → User

The orchestrator acts as a traffic controller, examining each request and routing it to the best agent for the task. Agents have access to tools, memory, and configuration specific to their role.

Built-in Agents

Agent	Purpose
Personal Assistant	General conversation, questions, planning
Coder	Programming tasks via Claude Code integration
Orchestrator	Routes requests to other agents

You can create unlimited additional agents through conversation or configuration files.

Next Steps

Ready to get started?

• Quick Start — Get Maestro running in minutes
• Architecture — Understand how the pieces fit together
• Creating Agents — Build your first custom agent