Active filters:
All
Loading video player...
Code - https://github.com/campusx-official/chatbot-in-langgraph https://github.com/campusx-official/mcp-client-langgraph In this video we continue the LangGraph Agentic AI playlist and implement MCP (Model Context Protocol) in LangGraph. First we recap what we’ve covered so far (Agentic AI vs Generative AI, LangGraph core concepts, workflows, and the chatbot project). Then we convert a synchronous chatbot to asynchronous (required by the MCP client library), build an MCP client in LangGraph, and connect it to a local math MCP server and a remote expense-tracker MCP server. The demo shows how MCP removes the maintenance brittleness of tool-based integrations by separating server-side tool logic from a lightweight client-side config. ============================ Did you like my teaching style? Check my affordable mentorship program at : https://learnwith.campusx.in DSMP FAQ: https://docs.google.com/document/d/1OsMe9jGHoZS67FH8TdIzcUaDWuu5RAbCbBKk2cNq6Dk/edit#heading=h.gvv0r2jo3vjw ============================ 📱 Grow with us: CampusX' LinkedIn: https://www.linkedin.com/company/campusx-official CampusX on Instagram for daily tips: https://www.instagram.com/campusx.official My LinkedIn: https://www.linkedin.com/in/nitish-singh-03412789 Discord: https://discord.gg/PsWu8R87Z8 E-mail us at support@campusx.in ⌚Chapters⌚ 00:00 – Intro and why the playlist was paused 00:56 – Recap of playlist (18 videos so far) 02:30 – Topic: what MCP is and why it matters 04:22 – Example: multiple tools (search, calculator, stock) and GitHub integration 07:48 – Problem: brittleness of tool-based approach and API version changes 11:31 – MCP solution: separation of concerns (server tools vs client config) 14:37 – Why we must convert sync → async (MCP libraries are async-only) 20:06 – Async conversion demo and a quick primer on async programming 26:35 – Building an MCP client in LangGraph (local math server demo) 33:24 – Adding a remote MCP server (expense-tracker) and listing tools 36:51 – Using expense-tracker: add, list, summarise — no client-side code changes needed 41:11 – Code structure and key changes (LangGraph, MCP client, Streamlit → async streams, SQLite → aiosqlite) 44:09 – Next video preview: add RAG capability