AI Engineering Handbook¶

The open-source path from zero to production AI

One sequential curriculum — transformers, RAG, agents, harnesses, evals, and LLMOps. No paywall. No module codes. Just start at course 01 and build real systems.

Start Here Browse all 16 courses ★ Star on GitHub

16 courses 140+ lessons 2 agent tracks MIT license

Who is this for?¶

New to AI

Software engineer or student starting from scratch

Start Here →

Know ML, need LLMs

ML practitioner catching up on transformers and APIs

Jump to Learn →

Building agents

Engineer shipping autonomous AI systems

Agent Engineering →

Using Claude Code / Cursor

Skills, loops, and context for IDE agents

Modern AI (2026) →

Shipping to production

Need LLMOps, evals, monitoring, safety

LLMOps & Production →

How to navigate¶

Start Here Learn Study Plans Build These Topic Map FAQ Setup

Goal	Go to
Follow the curriculum	Learn — 16 courses in order
New here	Start Here
Week-by-week schedule	Study Plans
Build a portfolio	Build These First
Find any topic	Topic Map
Questions	FAQ

The learning path¶

Part	Courses	Topics
Understand AI	01–05	NLP → neural nets → transformers → LLMs
Build applications	06–11	RAG, agents, harness, multi-agent, vector DBs, prompts
Production	12–14	LLMOps, evals, safety
Advanced	15–16	Fine-tuning, capstone projects

Optional tracks: Agent Engineering · Modern AI (2026)

Full course list →

Why this handbook?¶

	Typical blog / course	This handbook
Structure	Scattered posts	One sequential path, 16 courses
Depth	Surface-level	Engineer-grade foundations + production
Agents	Tutorial-only	Harness, MCP, multi-agent, evals
Cost	Paywalled	Free, MIT, forever

Contribute & star¶

If this helps you learn or ship AI systems, star the repo on GitHub — it helps others find it.

Improve a lesson, fix a link, or add an exercise: Contribute · Roadmap

GitHub →