Multi-Agent System
Definition
An architecture in which multiple specialized AI agents coordinate, each handling part of a task, to achieve a larger goal together.In-Depth Explanation
Instead of one monolithic agent, a multi-agent system divides work among specialized agents — for example a planner, a researcher, and a writer — that communicate and share context. An orchestration layer manages their workflow, role assignment, and conflict resolution. Frameworks like LangChain, CrewAI, and AutoGen make these systems easier to build. The approach can improve reliability on complex tasks but adds coordination overhead.
Real-World Example
A research assistant splits a report into sub-agents that gather sources, analyze data, and draft sections, then merges the results.