Apr 30
Loading video player...
Note: This video was made with an AI and scripting workflow as part of my build-in-public journey. Please let me know if you spot any inaccuracies or discrepancies so I can continue improving the process over time. By the end, you’ll understand how LangChain’s core packages — from LCEL pipelines to LangGraph state machines — assemble into a complete RAG application. That insight lets you choose the right components and patterns to build reliable, maintainable AI agents and retrieval pipelines. What you’ll learn: - How the LCEL pipe operator and Runnable primitives form composable AI pipelines - The role of LangGraph state graphs for branching, interrupts, and persistence - Configuring document loaders, text splitters, vector stores, and retrievers for RAG - Structuring prompts, models, embeddings, and output parsers for typed results - Defining and binding tools to extend agents with external functions - Setting up LangSmith tracing via environment variables and the shift from LangServe Chapters: CHAPTERS_PLACEHOLDER This video is for engineers and developers building AI-assisted applications who need a clear, code-focused introduction to LangChain’s architecture. Familiarity with Python and basic language-model APIs is required. Chapters 0:00 Langchain What It Is And How To Use It 0:52 Five Packages, One Story 1:55 Why Most Of LangChain Is Now LCEL 2:53 The Three Composition Primitives 3:51 Models, Embeddings, And Prompt Templates 4:51 Output Parsers And Structured Output 5:50 Tools And Tool Calling 6:44 Document Loaders And Text Splitters 7:45 Vector Stores And Retrievers 8:50 RAG, End To End 9:50 Why AgentExecutor Is Dead, And LangGraph Took Over 10:47 State Graphs, Nodes, And Edges 11:45 Interrupts And Human In The Loop 12:43 Threads And Persistence 13:38 LangSmith In Three Environment Variables 14:35 What Happened To LangServe 15:44 Five Use Cases A Five Person Company Can Ship This Week 16:53 The Interview Questions People Actually Ask 17:47 More Interview Questions, Going Deeper 18:54 When LangChain Is The Wrong Call 19:49 Your Five Day Learning Path 20:45 Carry These Five Ideas Forward