The AI Engineering Playbook
A zero-to-100 learning path for the consumer side of AI engineering — where you take models you did not train and build systems that are correct, fast, cheap, and safe enough to put in front of real users.
- 11
- sections
- 58
- pages
- 51
- diagrams
- 5
- worked designs
- 240
- quiz questions
- 168
- flashcards
The premise
Definitions are not understanding
Most material on this subject is a pile of definitions. Definitions tell you what something is called. They don't tell you what breaks without it, what it costs, or what you'd reach for instead — and those are the things you actually need when you're the one building it.
Why does this exist?
The specific pain that makes the concept necessary, and why the obvious alternatives lose.
How does it actually work?
One level down, with a diagram: the mechanism, the parameters, and the maths that matter.
What goes wrong in production?
Scale, cost, latency, failure modes — the things that only show up under real traffic.
Every page answers those three, in that order — then opens with its prerequisites and closes with where this connects, so the site is a web you can walk rather than a list you read front to back.
Fit
Who it's for
Written for you if
- You build at the application layer and want to understand these systems properly rather than assemble them from tutorials.
- You can code, and you've called an LLM API at least once.
- You'd rather know why a design loses than collect more of them.
Not what this is
- Not a machine-learning course. No PyTorch, no training loops, no backprop — model internals appear only at survival depth.
- Not a substitute for building something. Shipping a system is what makes it real.
- Not evergreen in its specifics. Prices and model names rot within months; mechanisms and tradeoffs don't.
The map
Eleven sections, in dependency order
Each section assumes the ones above it, and pages inside a section are ordered so nothing is used before it's introduced.
- 01LLM FundamentalsThe floor everything else stands on: what the model actually is, what it costs, what it can't do.8 pages
- 02Embeddings & SearchTurning meaning into geometry, then searching that geometry fast enough to matter.7 pages
- 03Vector DatabasesWhere the vectors live — and everything that gets hard once they do.4 pages
- 04RAGThe most-built system in applied AI — and the one most often described too shallowly.7 pages
- 05AgentsWhat happens when the model — not your code — decides what happens next.9 pages
- 06MCPThe Model Context Protocol — a standard socket between models and the tools they call.3 pages
- 07ProductionAnyone can build a demo. This section is everything between a demo and a system.7 pages
- 08Model LandscapeBuy, host, or adapt — and how to defend the choice.4 pages
- 09MultimodalWhen the input isn't text — and the moment your RAG pipeline meets a PDF full of charts.2 pages
- 10System DesignEverything above, assembled into one system, under constraints that conflict.7 pages
- 11Rapid FireCompression, not learning. Use this after you understand the material, not instead of understanding it.11 pages
The method
Built to be studied, not skimmed
Recognising an answer feels almost identical to being able to produce one, and the two come apart the moment you have to explain something out loud. So the site makes you produce it.
240 quiz questions
Scenario questions at the end of every page. Answer out loud, reveal, then grade yourself honestly — the grade is what counts, not the recognition.
168 flashcards
The rapid-fire decks in drill mode, with shuffle and a "weak only" filter that comes straight back to what you flagged.
Progress that follows you
Mark pages understood and watch the path fill in. Stored in your browser — no account, nothing sent anywhere.
Your copy
Where you are
Pick up where you left off, or start at the top.