Free Course 'AI Engineering from Scratch': 523 Lessons from Math to...

Free Course 'AI Engineering from Scratch': 523 Lessons from Math to Agents

Hacker News is discussing an open-source course 'AI Engineering from Scratch' (MIT license) by engineer Rohit Ghumare: 523 lessons in 20 phases, where every algorithm — from linear algebra and backprop to agents and swarms — is first derived 'from raw math,' and only then are frameworks like PyTorch introduced.

Compiled by Sergey KostenchukPublished 2026-10-03Updated 2026-10-03

2026-10-03 Coding Anthropic

💻 Free Course 'AI Engineering from Scratch': 523 Lessons from Math to Agents

Engineer Rohit Ghumare (rohitg00) published a project under the MIT license: 523 lessons in 20 phases, where every algorithm — from linear algebra and backprop to agents — is first derived from math, and only then is PyTorch introduced. Examples are written in Python, TypeScript, Rust, and Julia.

🌍 This is a shift from API-wrapper guides for LLMs to a full stack: attention, tokenizers, agent loops, MCP, production. The repository has gathered 62,904 stars on GitHub.

👤 Learning is free, no registration required: start with the first lesson or launch an AI tutor with the command npx skills add rohitg00/ai-engineering-from-scratch — interactive lessons in the terminal with a placement test. The six-volume EPUB/PDF is available in GitHub releases.

Source 1: https://aiengineeringfromscratch.com/index.html Source 2: https://github.com/rohitg00/ai-engineering-from-scratch

Sources