A hands-on crash course on LangGraph, the agent orchestration framework in the LangChain ecosystem. Covers building stateful agent workflows modeled as graphs, including: defining nodes and edges, custom state management, conditional routing based on intent classification, in-memory checkpointing for conversation context, RAG with vector stores, running Claude Code in headless mode, and implementing human-in-the-loop approval with looping revision flows. The tutorial builds a multi-agent system that classifies user intent (chat, knowledge, coding) and routes to specialized agents accordingly.
•53m watch time
Sort: