Multi-Agent System
Definition
Multiple agents work on a task with clearly defined roles.
A multi-agent system consists of several acting components that communicate or coordinate. They may use the same or different models and tools.
Simply explained
A team distributes research, calculation, and summarization among different members.
Why it matters in practice
Division of labor can structure complex processes and parallelize tasks.
Example from business
Hypothetical example: One agent collects product data, another creates a verified draft from it.
Limits and pitfalls
- More agents create additional coordination effort and possible chains of errors.
Origin & history
Multi-agent systems originate from distributed AI; language models expand their application possibilities.
Use in mid-sized companies
- SMEs should define clear handovers and responsibilities before deciding on the number of agents.
Frequently asked questions
Does each task need multiple agents?
No, simple processes are often better handled with a workflow.
Sources
- Multi-agent orchestration — · OpenAI Agents SDK: Multi-agent orchestration —
Editorially reviewed on: