Loading video player...
Learn how to add memory to your agent → https://mdb.link/lLrys76gPlM-ai-memory Learn how to elevate AI applications by building sophisticated, stateful agents using the LangGraph framework and MongoDB. This tutorial moves beyond simple query-response chains, demonstrating how to implement multi-step reasoning and cyclical workflows that allow your agent to make decisions, use tools, and maintain context across complex interactions. In this lesson, discover the core components of a LangGraph application, including nodes, edges, and state management. Follow along to build a functional MongoDB agent capable of performing vector searches and direct page lookups to answer technical documentation queries with high accuracy. See what Atlas is capable of for free: https://mdb.link/YT-Atlas-Register 00:00:00 Introduction to LangGraph and Advanced Agents 00:00:54 The Power of Graph Data Structures for AI 00:01:33 Understanding LangGraph: Cycles and State 00:02:44 Defining the Graph State and Reducers 00:04:11 Building the Agent Node: The Reasoning Engine 00:05:18 Creating the Tool Node: Executing Actions 00:06:06 Implementing Conditional Routing and Logic 00:07:48 Assembling the Complete State Graph 00:08:44 Executing the Graph and Streaming Results 00:10:01 Demo: Testing the MongoDB Documentation Agent 00:11:11 Final Summary and Key Takeaways Resources: MongoDB main YouTube channel: https://www.youtube.com/@MongoDB Website: https://mdb.link/MongoDBYT LinkedIn: https://www.linkedin.com/company/mongodbinc MongoDB Developer Blog: https://mdb.link/developerblogYT