Building Agentic and Multi-Agent Systems with LangGraph
summary
Claims that LangGraph turns LangChain-style flows into graph-based cyclical loops that make adding agents and multi-agent teams simple, and demonstrates concrete code: Python nodes, structured tools (weather API + Duck/Tavily search), an OpenAI function-calling model (GPT family) wired to a tool belt, and stateful subgraphs. The demo builds a meteorologist/research/reporting multi-agent pipeline that writes a Toronto weather report to disk, traces execution via LangSmith, and debugs with the LangGraph Studio agent IDE (Mac only), showing the approach works end-to-end.
AI-written summary · about
chapters (13)
- 0:00Introduction to AI Makerspace
- 2:22Core Generative AI patterns
- 5:11Prompting and RAG mechanics
- 7:42Evolution of LLM reasoning
- 15:51Introduction to AI agents
- 19:45Reasoning and action patterns
- 24:28LangGraph overview
- 30:41Practical agent implementation
- 49:09Multi-agent systems theory
- 55:49Orchestration framework comparison
- 1:14:48Building complex agent teams
- 1:39:01LangGraph Studio IDE
- 1:49:42Final thoughts and Q&A
related talks
25mFrom LangChain to LangSmith and LangGraph: the Future of AI
01h 03mAgents Everywhere
53mHands-on with AI agents: From scratch to LangGraph
56mOrchestrating Agentic State Machines with LangGraph
28mAI Agents Graph: Your Following Tool in Your Java AI Journey
29mBuilding Modular AI Agents with LangGraph, MCP, and
01h 27mBlazing the AI Trail: Using LangGraph to Conquer the Oregon Trail
29mTransparent Multi-Agent AI