Multi-Agent Systems with LangGraph

Overview

Build powerful multi-agent systems by applying emerging agentic design patterns in the LangGraph framework. Build AI Agents with LangGraphDesign and build your own agents with LangGraph!

LangGraph is a core part of the LangChain ecosystem, and it's used to build production-ready AI agents with a high degree of customizability. LangGraph allows developers to build agents as graphs with nodes and edges, which allows information flow and decision pathways to be carefully mapped out, and reduces the room for unexpected errors to creep in.Explore Emerging Multi-Agent ArchitecturesSince the rise of AI agents, a handful of agentic design patterns have emerged, and you'll learn about two of the most popular:

network (or decentralized) multi-agents and supervisor multi-agents.

Manage multiple agents effectively by designing a supervisor agent to delegate tasks and encourage collaboration between the worker agents.Create Your Own Agentic AssistantYou'll use LangGraph to build an agentic assistant to gather information and stock performance data on Fortune 500 companies, and analyze it using visualizations!You'll see this agent progress from a simple single-agent system to a three-agent supervisor multi-agent! Join the growing number of AI builders and learn to design and build AI agents today!

Syllabus

• Agents as Graphs
Learn to build AI agents the LangGraph way! Build a toolbox of tools to help your agent interact with APIs, retrieve data from CSV files, and run Python code! Begin to build a single-agent system using nodes and edges to connect the LLM and tools in a controlled and methodical way.
• LangGraph Multi-Agents
Since the rise of AI agents, a handful of agentic design patterns have emerged, and you'll learn about two of the most popular: swarm (or decentralized) multi-agents and supervisor multi-agents. You'll see that LangGraph provides a whole host of functionality to design multi-agents tailored to your specific use case.

---

Taught by

James Chapman

---

Subjects

Computer Science