Multi-Agent System
An AI architecture where multiple specialized agents collaborate, debate, or coordinate to solve complex problems that are difficult for a single agent.
Patterns
Supervisor: One agent coordinates others. Debate: Agents argue for different answers, improving accuracy. Pipeline: Agents process tasks sequentially (researcher -> writer -> editor). Swarm: Agents dynamically hand off tasks based on expertise.
Benefits
Specialization (each agent optimized for a role), error checking (agents review each other's work), scalability (add more agents as needed), and handling complex workflows that require multiple capabilities.
Frameworks
CrewAI, AutoGen (Microsoft), LangGraph, Claude Agent SDK, and Swarm (OpenAI). Multi-agent systems are a rapidly evolving area of AI engineering.