What is Multi-Agent System?
An architecture where multiple specialized AI agents collaborate to complete complex tasks.
Definition
A multi-agent system is an AI architecture where multiple agents — each specialized for a specific role or task — work together to accomplish goals that would be too complex or slow for a single agent. Agents communicate by passing outputs to each other, with an orchestrator managing the overall flow. Common patterns include supervisor agents that delegate to worker agents, and parallel agents that handle different subtasks simultaneously.
Example
A content production system: a research agent finds relevant information → a writing agent drafts the article → an SEO agent optimizes it → a publishing agent posts it. Each agent is specialized; together they complete the full workflow.
Multi-Agent System vs ai-orchestration: What's the difference?
An architecture where multiple specialized AI agents collaborate to complete complex tasks.
Multi-agent system describes the architecture (multiple agents). Orchestration is the mechanism that coordinates them.