Home › Courses › Multi-Agent Systems & Swarms

Multi-Agent Systems & Swarms

When a second agent earns its cost, which topology to use, how agents reach agreement, and how multi-agent systems fail

A systems course on more than one agent. It starts from why multi-agent at all and the vocabulary that predates LLMs — FIPA-ACL, speech acts, communication protocols — then works through the topologies and how each one breaks: supervisor/orchestrator-worker, hierarchies, society-of-mind debate, role specialization, parallel swarms, group chat and stateless handoffs. The middle is coordination: the A2A protocol, blackboards, consensus under Byzantine faults, voting and debate topology, negotiation, emergent simulation and theory of mind. It ends on production — swarm optimization, multi-agent RL, agent economies, queues and durability — and on measurement: the MAST failure taxonomy, groupthink and cascading errors, and coordination benchmarks.

4 phases · 25 lessons · 25 worked examples · Early access

Take Agentic AI Engineering first — this course builds on it.

Every phase, every lesson, every project

• Why Multi-Agent, and the Primitives (4 lessons) — free — When a second agent earns its cost, and the vocabulary that predates LLMs: FIPA-ACL and speech acts, communication protocols, and the primitive model every framework is a variation on.
  • Why Multi-Agent?
  • Heritage of FIPA-ACL and Speech Acts
  • Communication Protocols
  • The Multi-Agent Primitive Model
• Topologies & Roles (7 lessons) — The shapes a multi-agent system can take and how each one fails: supervisor/orchestrator-worker, hierarchies, society-of-mind debate, role specialization, parallel swarms, group chat with speaker selection, and stateless handoffs.
  • Supervisor / Orchestrator-Worker Pattern
  • Hierarchical Architecture and Its Failure Mode
  • Society of Mind and Multi-Agent Debate
  • Role Specialization — Planner, Critic, Executor, Verifier
  • Parallel / Swarm / Networked Architectures
  • Group Chat and Speaker Selection
  • Handoffs and Routines — Stateless Orchestration
• Coordination & Agreement (7 lessons) — Getting independent agents to agree on something: the A2A protocol, blackboards and shared memory, consensus under Byzantine faults, voting and debate topology, negotiation, emergent simulation, and theory of mind.
  • A2A — The Agent-to-Agent Protocol
  • Shared Memory and Blackboard Patterns
  • Consensus and Byzantine Fault Tolerance for Agents
  • Voting, Self-Consistency, and Debate Topology
  • Negotiation and Bargaining
  • Generative Agents and Emergent Simulation
  • Theory of Mind and Emergent Coordination
• Scale, Failure & Evaluation (7 lessons) — What breaks in production and how it is measured: swarm optimization, multi-agent RL, agent economies and reputation, queues and durability, the MAST failure taxonomy, groupthink and cascading errors, coordination benchmarks, and the 2026 state of the art.
  • Swarm Optimization for LLMs (PSO, ACO)
  • MARL — MADDPG, QMIX, MAPPO
  • Agent Economies, Token Incentives, Reputation
  • Production Scaling — Queues, Checkpoints, Durability
  • Failure Modes — MAST, Groupthink, Monoculture, Cascading Errors
  • Evaluation and Coordination Benchmarks
  • Case Studies and the 2026 State of the Art

The technologies you will use

A2A · LangGraph · AutoGen

Common questions

Does this course have graded labs?

Not yet. Each lesson ships the worked example it refers to and states its exercises, but nothing here is checked by the CLI the way the from-scratch courses are. The course is marked early access for that reason.

Key terms in this course

Agent · Multi-Agent System · Orchestration

Continue your learning path

Agentic AI Engineering · Build a Multi-Agent Research Team · Autonomous Agents & Frontier Safety

Start the Multi-Agent Systems & Swarms course

Create a free account — the opening phases of 29 of 35 courses are free, no credit card. Or see Pro pricing.

All courses · Pricing · About · FAQ · Glossary