AI Engineer Notes

A twelve-week path · 66 notes · 4 projects

AI Engineer Notes

From “what is a neural network?” to designing LLM systems in an interview. Every idea is explained simply first, then precisely, then in words you can say out loud.

These notes assume no AI background. Each one starts with an everyday explanation, shows the idea as a picture, then gives the exact technical terms, the math and the code. It ends with a model interview answer, practice questions and a revision card.

Start with the roadmap to see what to study each week, then check the projects you will build alongside. 1 of 66 notes are written so far.

W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 W11 W12 study 1 Basics 2 LLM internals 3 Modern 4 RAG 5 Agents 6 Prod 7 Interview drill build P1 GPT from scratch P2 LoRA P3 RAG + evals P4 Agent + MCP, production layer Polish + resume
Fig. 1. The twelve weeks at a glance: what you study (top) and what you build (bottom).

The seven phases

PhaseWhenNotes
1. Deep learning foundations Week 1 7
2. LLM internals: build GPT from scratch Weeks 2–4 18
3. The modern LLM stack Week 5 8
4. Retrieval-augmented generation (RAG) Weeks 6–7 9
5. Prompting, tools and agents Weeks 8–9 9
6. Evals and production Week 10 7
7. LLM system design and interview drill Weeks 11–12 8