Credits

Standing on the shoulders of people who shared their work. Thank you.

The course

AI Engineering from Scratch by Rohit Ghumare and contributors, MIT licence. The AI tracks and Exam prep follow its lessons; we rewrote the teaching in our own words and built new 3D scenes. Code shown on lesson pages is adapted from it under the same MIT licence.

Build it yourself

The tutorials come from the list build-your-own-x. Each lesson names its tutorial and author, and keeps that tutorial's licence for anything adapted from it.

Build your own database: DBDB: Dog Bed Database by Taavi Burns, in 500 Lines or Less (text CC BY 3.0, code MIT). Our lesson retells the idea; the code shown is our own short version.

Software and fonts

What	Used for	Licence
Three.js (r160)	The 3D scenes	MIT
Kokoro (voice Heart)	The narration, recorded once, ahead of time	Apache 2.0
PocketBase	Accounts and saved progress	MIT
Bricolage Grotesque	Headings	SIL Open Font Licence
Instrument Sans	Text	SIL Open Font Licence
JetBrains Mono	Code	SIL Open Font Licence

Lesson by lesson

#	Lesson	Built from
1	Your dev environment	01-dev-environment
2	Git: save every version	02-git-and-collaboration
3	GPUs: why AI needs graphics cards	03-gpu-setup-and-cloud
4	APIs and keys	04-apis-and-keys
5	Notebooks and the kernel	05-jupyter-notebooks
6	Python environments and lock files	06-python-environments
7	Docker: the same box everywhere	07-docker-for-ai
8	An editor that helps	08-editor-setup
9	Data: splits, formats and streaming	09-data-management
10	The terminal: pipes and sessions	10-terminal-and-shell
11	Vectors: how alike are two things?	02-vectors-matrices-operations
12	Gradient descent	08-optimization
13	Linear regression	02-linear-regression
14	The perceptron	01-the-perceptron
15	Backpropagation	03-backpropagation
16	Activation functions	04-activation-functions
17	Convolutions	02-convolutions-from-scratch
18	Tokens	02-building-a-tokenizer
19	Word embeddings	03-word-embeddings-word2vec
20	Self-attention	02-self-attention-from-scratch
21	Multi-head attention	03-multi-head-attention
22	Positional encoding	04-positional-encoding
23	Next word and temperature	07-gpt-causal-language-modeling
24	Quantization	11-quantization
25	RAG: look it up first	06-rag
26	Function calling	09-function-calling
27	The agent loop	01-the-agent-loop
28	MCP basics	06-mcp-fundamentals
29	Build your own database	A build-your-own-x tutorial (credited on the lesson page)
30	Exam prep: designing tools	18-tool-contracts-errors-and-progressive-discovery

Made with AI

This site is made with AI, and we say so plainly. Claude, an AI model made by Anthropic, wrote the code, the 3D scenes and the first drafts of the lessons, working to our design and direction. People chose what to teach and how it should look and feel, and every lesson is reviewed before it goes up. Using Claude to build the site does not mean Anthropic endorses it.

If you spot a mistake, tell us: AI and people both get things wrong, and we fix errors as soon as we hear about them.