Microsoft · GitHub
AI for Beginners
24 lessons for the foundations: AI concepts, neural networks, computer vision, NLP, and responsible AI.
Best first step if you are completely new to AI.
From the reel · Free learning path
Want to build AI agents?
Start here.
These are the 10 courses from my Reel, in the same order. Start with the foundations, then move into MCP, LangChain, LangGraph, and building your own agent.
The simple route
- Start with AI and generative-AI basics.
- Learn MCP so your agent can use tools and data.
- Build with LangChain and LangGraph.
- Ship one agent and learn by improving it.
The complete list
10 free courses, linked.
All links go to the official course or repository.
Open course ↗
Microsoft · GitHub
Generative AI for Beginners
21 lessons on building generative-AI applications, with written lessons and Python or TypeScript examples.
Take this after the basics if you want to build with LLMs.
Microsoft · GitHub
AI Agents for Beginners
18 lessons on agent design, tools, RAG, planning, multi-agent patterns, memory, and production concerns.
A solid overview before choosing a framework.
Microsoft · GitHub
MCP for Beginners
Learn the Model Context Protocol, then build a first server and client that connect agents to tools and data.
Good if you want to understand what MCP actually connects.
Hugging Face · Official course
Hugging Face MCP Course
A free course built with Anthropic that covers MCP fundamentals and hands-on server and client work.
The course uses free learning material; projects may still need services or API credentials.
LangChain Academy · Official course
LangChain Essentials
18 lessons on creating an agent, tools, MCP, memory, structured output, and human approval.
Start here when you want a fast hands-on LangChain introduction.
LangChain Academy · Official course
LangGraph Essentials
13 lessons on nodes, edges, memory, conditional paths, and human-in-the-loop workflows.
Use it to understand the building blocks before the longer course.
LangChain Academy · Official course
Introduction to LangChain
30 lessons that go from a first agent to MCP, multi-agent systems, long conversations, and human review.
Take it once you are ready to build a more complete agent.
LangChain Academy · Official course
Introduction to LangGraph
55 lessons on state, memory, human feedback, subgraphs, long-term memory, and deployment.
This is the deeper course for durable, more controllable agent workflows.
Hugging Face · Official course
Hugging Face Agents Course
Learn agent fundamentals and frameworks, then build, test, and certify your own agent in the final project.
The best final project if you want proof of work, not only notes.
Do this first
Pick one small agent to build.
Try a personal research helper: give it one search tool, one source of notes, and a rule that it must show the links it used. Do not try to build a ten-tool “autonomous” agent on day one.
FIRST PROJECT BRIEF
Help me build a small research agent. It should search one approved source, save a short note with links, and ask me before taking any action. Start with a simple plan, then build the smallest working version.
Before you start
Every course above is free to access when checked on 22 September 2026. Some code examples can require an account, a local development setup, or a paid model/API if you choose to use one. Read each course’s current setup guide before you run its examples.