Multi-agent system
A multi-agent system is a computational system composed of multiple autonomous or semi-autonomous agents that interact or coordinate to perform tasks or pursue goals.
Build a Multi-Agent System (from Scratch)
A multi-agent system is a system composed of multiple interacting agents that cooperate, coordinate, or compete to achieve goals.
- How it is used
- Coordinates specialized LLM agents on decomposable tasks and combines their results.
- How deep it goes
- Explains subagent registry and dispatch flow, while deferring detailed failure modes and remedies.
- Examples
- Uses a coordinator delegating New York food searches to croissant, pizza, and remote bagel agents.
- What you can do after
- Build it from scratch using subagent specifications, a registry, and a dispatch tool.
Read how Build a Multi-Agent System (from Scratch) explains multi-agent system →
Introduction to Generative AI, Second Edition
A multi-agent architecture is a system design in which multiple software agents coordinate or interact to perform tasks.
- How it is used
- Uses coordinated agents to divide complex workflows into specialized subtasks and pursue shared goals.
- How deep it goes
- Explains roles, handoffs, shared context, and collective risks at an overview level without an implementation walkthrough.
- Examples
- Illustrates coordination with employee onboarding, specialized research roles, and drug-discovery agents.
- What you can do after
- Evaluate/understand how agent coordination can support complex tasks while introducing collective failures that individual-agent tests may miss.
Read how Introduction to Generative AI, Second Edition explains multi-agent system →
AI Agents in Action
A multi-agent system is a system in which multiple specialized agents work and communicate together to solve problems.
- How it is used
- Coordinates specialized agents to tackle complex tasks and exchange feedback.
- How deep it goes
- Explains communication patterns and walks through setup, configuration, and execution of a basic agent chat.
- Examples
- Uses a coder and tester, a proxy-led coding task, and a data-fetching, analysis, and reporting exercise.
- What you can do after
- Apply it via AutoGen or CrewAI to coordinate agents on coding, summarization, or reporting tasks.
AI Agents in Action, Second Edition
A multi-agent system is a system in which multiple autonomous or semi-autonomous agents coordinate to perform tasks or achieve shared or related goals.
- How it is used
- Extends single-agent systems for specialized, parallel, context-distributed, or inherently multi-actor tasks.
- How deep it goes
- Explains coordination patterns, decision authority, control, and communication, while treating enterprise security, governance, compliance, and scaling briefly.
- Examples
- Uses customer-support specialists and research agents that select, fetch, and summarize sources.
- What you can do after
- Evaluate and understand coordination patterns by comparing agent roles, decision authority, communication, and trade-offs.
Read how AI Agents in Action, Second Edition explains multi-agent system →
Vibe Engineering
A multi-agent system is a software system in which multiple autonomous or semi-autonomous agents operate concurrently or collaboratively toward specified objectives.
- How it is used
- Coordinates parallel coding agents while a developer reviews and merges their work.
- How deep it goes
- Explains task decomposition, workspace isolation, cost tradeoffs, and specification risks, without providing an implementation walkthrough.
- Examples
- Uses an API/frontend/test split, a measured two-agent conflict rate, and a concurrent Rust browser-building experiment.
- What you can do after
- Evaluate multi-agent coding work against sequential execution using task independence, coordination costs, and specification clarity.
Building Reliable AI Systems
A system composed of multiple AI agents that coordinate to perform tasks.
- How it is used
- Coordinates specialized agents to handle multipart customer requests that combine product search, policy questions, and images.
- How deep it goes
- Builds specialist agents and a stateful orchestrator, then compares explicit graph routing with role-based coordination.
- Examples
- Uses a shopping assistant to combine product search, policy retrieval, and vision-based search for customer queries.
- What you can do after
- Apply it via LangGraph to route requests among specialists, share state, and synthesize responses.
Read how Building Reliable AI Systems explains multi-agent system →
The Art of AI Product Development
A multi-agent system is an AI system in which multiple agents perform specialized tasks and coordinate toward a shared goal.
- How it is used
- Coordinates specialized agents to address complex, dynamic business workflows that exceed a single agent’s capabilities.
- How deep it goes
- Explains specialization, orchestration, communication, and supervision, but stops short of implementation details.
- Examples
- Uses product discovery, prioritization, roadmap planning, customer-feedback analysis, and future-scenario generation as collaborative tasks.
- What you can do after
- Operate/configure specialized agents incrementally, monitor gaps, and add human review when confidence is low or errors could be catastrophic.
Read how The Art of AI Product Development explains multi-agent system →
Grokking AI Applications
An AI system composed of multiple agents that collaborate or coordinate to accomplish tasks.
- How it is used
- Organizes complex AI workflows through specialized agents that collaborate, delegate, or execute in sequence.
- How deep it goes
- Explains motivations, benefits, and sequential or hierarchical structures, while deferring detailed implementation to iterative design and testing.
- Examples
- Uses planner, calculator, fact-checker, billing, support, documentation, scheduling, budgeting, and compliance agents.
- What you can do after
- Designs and evaluates role-specific agent workflows with controlled delegation, communication, and iterative testing.
Read how Grokking AI Applications explains multi-agent system →
AI Governance
A multi-agent system is a system in which multiple agents interact or collaborate to perform tasks.
- How it is used
- Uses collaborating agents to solve tasks within agentic workflows.
- How deep it goes
- Defines agent collaboration and gives a code example, while noting orchestration complexity and failure paths.
- Examples
- Illustrates code generation paired with review or testing, and IT support agents that run diagnostics and collaborate.
- What you can do after
- Understand how agent collaboration can catch mistakes while adding orchestration complexity and failure paths.
Investing for Programmers
A multi-agent pattern is an architecture in which multiple specialized agents collaborate through coordinated and integrated reasoning.
- How it is used
- Coordinates specialized agents when complex analysis benefits from parallel work, delegated reasoning, and integrated judgments.
- How deep it goes
- Explains specialization, parallelism, orchestration, and deliberation, but does not specify implementation protocols or agent code.
- Examples
- Uses an investment workflow and a courtroom-style panel of optimist, skeptic, and adjudicator agents.
- What you can do after
- Evaluate multi-agent investment-analysis designs by mapping specialized roles to coordinated recommendations.
Read how Investing for Programmers explains multi-agent system →
Build an AI Agent (From Scratch)
A multi-agent setup is an architecture in which multiple agents cooperate to complete a task.
- How it is used
- Coordinates specialists when a single agent’s growing responsibilities make prompts and tool selection harder to manage.
- How deep it goes
- Explains three collaboration patterns through control, context flow, interfaces, and composable execution structures.
- Examples
- Uses a technical blog assistant with research, coding, writing, and review specialists.
- What you can do after
- Operate/configure a multi-agent system by assigning specialist roles and combining workflow, tool, and transfer patterns.
Read how Build an AI Agent (From Scratch) explains multi-agent system →
AI Agents and Applications
A multi-agent system is a software system in which multiple agents independently or cooperatively perform tasks and pursue goals.
- How it is used
- Coordinates specialized agents to answer domain-specific requests, perform actions, and combine results across complex workflows.
- How deep it goes
- Explains router and supervisor architectures, graph assembly, communication, tracing, and reliability practices rather than agent internals.
- Examples
- Uses travel information, accommodation booking, weather, hotel availability, customer service, research, and trading workflows as examples.
- What you can do after
- Build it from scratch with LangGraph using router or supervisor patterns, specialist tools, structured state, and tracing.
Read how AI Agents and Applications explains multi-agent system →
Designing AI Agents
A software system in which multiple agents coordinate or work independently to perform tasks.
- How it is used
- Uses multiple agents when delegation supplies capabilities or parallel work that justify coordination costs.
- How deep it goes
- Explains coordination costs, failure modes, and primary-agent ownership while deferring detailed handoff, fan-out, and hierarchy patterns.
- Examples
- Illustrates parallel company research and compares results on sequential-planning and decomposable financial-reasoning benchmarks.
- What you can do after
- Evaluate whether to add agents using task success, latency, token cost, and debugging effort.
Also covered in
- AI Agents for Offensive SecurityIn early access (MEAP)
- Knowledge Graphs and LLMs in Action
- Architecting Autonomous Agentic SystemsIn early access (MEAP)
- Context EngineeringIn early access (MEAP)
- Build Applications with Local AI Models on a Mac
- Agent Design PatternsIn early access (MEAP)
- Digital Twins in Action
What to know first
- language modelA language model is a computational model that represents patterns in linguistic sequences to assign probabilities, predict, generate, or evaluate language.
- AI agentAn AI agent is a computational system that uses artificial intelligence to interpret information and select or perform actions toward goals while interacting with an environment.
- toolA software tool is a capability or utility that enables a person or system to perform a defined task or operation.
- agent orchestrationAgent orchestration is the coordination of agents, their tasks, tools, and interactions to accomplish a larger objective or workflow.
Related concepts
- LangGraphused by · LangGraph is a framework for building stateful, graph-structured workflows and applications, including those involving language models and agents.
- LLM-powered agenthas part · An LLM-powered agent is a software agent that uses a large language model to interpret inputs and generate responses or actions toward a goal.
- AI agentuses · An AI agent is a computational system that uses artificial intelligence to interpret information and select or perform actions toward goals while interacting with an environment.
- specialized agenthas part · A specialized agent is an agent configured to focus on a particular domain or class of tasks, with responsibilities or capabilities suited to that focus.
- subagenthas part · A subagent is an agent assigned by another agent to perform a delegated part of a larger task under its coordination or authority.
Questions about multi-agent system
What is multi-agent system?
A multi-agent system is a computational system composed of multiple autonomous or semi-autonomous agents that interact or coordinate to perform tasks or pursue goals.
Which Manning books explain multi-agent system?
Build a Multi-Agent System (from Scratch), Introduction to Generative AI, Second Edition, AI Agents in Action, AI Agents in Action, Second Edition, Vibe Engineering, Building Reliable AI Systems, The Art of AI Product Development, Grokking AI Applications and 5 more.
What should I know before learning multi-agent system?
Manning books that teach multi-agent system usually introduce language model, AI agent, tool and agent orchestration first.
What kinds of multi-agent system are there?
Kinds of multi-agent system include interoperable multi-agent system.
What are the parts of multi-agent system?
Multi-agent system includes LLM-powered agent, specialized agent and subagent.
What is multi-agent system also called?
Multi-agent system is also called multi-agent architecture, multiagent system, MAS or multi-agent pattern.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.












