From zero to AI engineer
Learn the fundamentals by watching, answering and running real code in the page. Then build four projects, ending with an AI copilot real people use.
- Lessons
- 104
- Hours of video
- 10
- Exercises you run
- 73
How LLMs actually work
Every chatbot, summarizer, and agent you build sits on one mechanism: next-token prediction.
Each lesson runs the same way: a short narrated video that stops to ask you a question once it has taught the answer, then code you write and run right in the page, in JavaScript, Python or SQL. Nothing to install.
Inside the course
Sixteen modules in seven parts. Start where you are; jump anywhere.
- Start hereSet up your machine, think like a programmer, write your first code.
- Web foundationsHow the web works, then HTML and CSS you can reason about.
- JavaScript and the frontendThe language properly, then React and TypeScript.
- Backend and dataNode and APIs, SQL, schema design and transactions.
- Security and shippingAuth and the attacks that matter, CI, containers and the cloud.
- AI and agentsLLMs from the API side, RAG, agents, MCP, and how models are made.
- The AI-native engineerTaste, specs and verification; Claude Code; Python as a second language.
- lessons with a narrated video
- 104
- quiz questions, asked mid-video
- 312
- coding exercises with checks
- 73
- portfolio projects
- 4
What you’ll build
- After web foundationsA personal site, hand-rolledHTML, CSS and a little JavaScript. No frameworks.
- After backend and dataTaskFlowA real full-stack app: React, an API, a database and sign-in.
- After security and shippingShip it like productionTests, CI, containers and a deploy you’d let strangers use.
- Graduation projectResearch CopilotAn AI agent over your own documents and tools, shipped to real users.
Your first lesson takes about seven minutes. Progress saves in your browser as you go.
Free, no signup, nothing to install
Start lesson 1