Progress: 0%
A Complete Practical Course

Building a Private LLM to Agent AI

Go from an empty Ubuntu machine to a fully private AI agent ecosystem — local LLMs, RAG, and multi-agent systems, running entirely on your own hardware. No cloud account required.

Start Module 0 Check Requirements First
24
Core Modules
10
Capstone Projects
6
Bonus Tracks
$0
Cloud Accounts Required

The Six-Part Roadmap

  1. Part 1
    Foundation
    Linux, Python, VS Code, Git, Docker, APIs.
  2. Part 2
    Local LLM
    Ollama, running real models, prompt engineering, fine-tuning.
  3. Part 3
    AI Applications
    LangChain, LangGraph, vector databases, RAG, evaluation, MCP.
  4. Part 4
    Agent AI
    Agents, tools, memory, multi-agent systems.
  5. Part 5
    Production Ready
    Databases, FastAPI, a frontend, Docker Compose, real deployment.
  6. Part 6
    Capstone Projects
    Ten complete products combining everything above.
No GPU? No problem.

Every hands-on exercise in this course has a CPU-only path. A GPU speeds things up but is never required to complete any module.

Where to Go Next

Start Here
Confirm your setup before Module 1.
Module 0
The full course introduction.
Capstone Projects
See what you'll eventually build.