Multi-Agent Systems
Some jobs are too big for one AI agent. A multi-agent system shares the job between several agents. Each agent has one clear job. This lesson shows the common shapes, the costs, and a simple rule for when to use them.
New words
- Multi-agent system
- Several AI agents that work together on one big job.
- Orchestrator (lead)
- The agent that splits the job, sends work and joins the results.
- Helper agent
- An agent that does one part of the job for the lead.
- Pipeline
- Agents in a line, where each one passes its result to the next.
- Brief
- A short written message that tells an agent what to do.
- Workflow
- A fixed list of steps written by a person, with no choices by the model.
What you will learn
- One volunteer, one big fest
- Why does one agent struggle?
- Three common shapes
- A clear job, and how work passes
- Try it: one task, three ways
- Costs and risks
- When one agent is better, and a simple rule
- In a real system
- Interview tip
- Key takeaways
10 quiz questions at the end.
The full lesson, with interactive demos and the quiz, opens after you sign in with Google. It is free.