AI Agents with LangGraph & MCP: Build RAG & Multi-Agents

About this course

This course contains the use of artificial intelligence.

AI agents are moving from demos into real products. This course shows you how to build them properly.

You will build AI agents in Python using LangGraph, the framework for stateful agent workflows, and connect them to real tools through the Model Context Protocol (MCP). We start with a simple tool-calling agent and build up, one step at a time, to retrieval-augmented (RAG) agents and multi-agent systems that work together on a task.

Every concept comes with code. You will see why an agent picks a tool, how state moves through a graph, how to keep the agent's answers grounded in your own documents, and how to split a large job across several specialised agents.

What you will buildA tool-calling agent that decides when to search, calculate or call an APIA RAG agent that answers questions from your own documents with retrieved context

An MCP server that exposes your own tools to any MCP-compatible clientA multi-agent system in which a supervisor routes work to specialist agents

How the course works

Short, focused video lectures with captions

Complete source code for every project

Quizzes at the end of each section to check your understanding

Hands-on assignments so you write the code yourself

Who this course is for

This course is for Python developers, data scientists and ML engineers who already know the basics of LLMs and want to build agents that do real work. If you have called an LLM API before and want to go further, this course is for you.

By the end, you will be able to design, build and debug your own agent workflows with LangGraph and MCP.

Prices, discounts and availability are set by Udemy. We may earn a commission when you purchase through links on this site.