From bedrock to frontier.
Master the AI-engineering skill map.
The most rigorous, interactive way to learn production AI, 4 tiers, 27 tracks, 210 chapters, and 18 real-world capstones, sourced from the labs that built the field. Not videos, real systems you run, in the language you ship in.
Run a real retriever, right now.
Type a question. A real TF-IDF retriever searches a knowledge base in your browser, scores every chunk by cosine similarity, and grounds its answer in what it finds, or refuses when the evidence is too weak. The same engine as the RAG chapter; ask something off-topic and watch it refuse instead of hallucinate.
Don't watch. Write the code, we'll run it for real.
This isn't a mock terminal. Your Python runs on CPython compiled to WebAssembly, your TypeScript is checked by the real tsc, your Java is compiled by the real javac and executed on a JVM, right here, in your browser. Fix the retry policy below and make the tests pass; every failure shows the full traceback, exactly like production. This is how every chapter teaches.
One climb. 4 tiers. Bedrock to frontier.
Place out of what you already know, go deep on what you don't. Start at the engineering and systems bedrock, ascend through the 30-chapter Applied AI core, and top out at the staff-level specializations, 27 tracks, 210 chapters, one continuous ladder.
Foundations
Live7 tracks · 65 chaptersThe beginner on-ramp, AI & LLM literacy, programming basics, the math behind models, and the developer toolkit. The free funnel that takes you from zero to ready (or place out with a quick test).
Systems
Live7 tracks · 59 chaptersThe engineering and systems bedrock behind real AI, Linux, networking, data systems, distributed systems, security, and GPU/inference. This is where you separate from shallow AI bootcamps.
Applied AI
Live6 tracks · 30 chaptersBuild production-grade AI systems, the 30-chapter Applied AI Engineering spine, every chapter a bespoke interactive lab. The flagship of the curriculum.
Mastery
Live7 tracks · 56 chaptersOperate AI at scale, the staff+/architect/enterprise specializations: infrastructure, system design, evaluation, cost engineering, product, debugging, and incident response.
Not a course you watch. A game you level up in.
Every question earns XP. Every section you certify lights up your skill tree. Streaks, spaced review, badges and a live leaderboard turn the long climb to staff-level into something you come back to — daily.
- XP & rising levels
- Certification exams per section
- Daily streaks & spaced review
- Mastery, badges & leaderboard
Today's quests
resets daily- Show up — open a lesson+10 XP
- Sharpen — review 5 cards+20 XP
- Certify — pass a section exam+50 XP
This week's leaderboard
season resets weekly- ada.builds2,140XP
- youyou1,980XP
- kenji_ml1,715XP
18 real-world capstone projects.
Each turns the chapters' mental models into a shipped, measured system on real infrastructure, every one with its own interactive architecture diagram and a step-by-step build guide. Finish six and you have the portfolio that ends the interview.
Production RAG Over Ugly Real Documents
A law / insurance / clinical firm hands you 500 genuinely ugly PDFs, scans, multi-column, tables, and demands grounded answers with citations and zero hallucination tolerance.
View build planA Multi-Agent System on a Real Framework
Automate a real knowledge-work task end-to-end: a competitive-research report, a codebase migration, or a due-diligence memo.
View build planArchitect & Defend a Real System Design
The actual staff/architect interview, made real: "Design an AI assistant over 10 years of our wiki + support tickets + code."
View build planSourced from the labs, tools & research that define the field.
Every chapter ends in numbered, URL-backed sources spanning 80+ organizations. You learn how production AI is actually built, straight from the people who built it.
One craft. Every timezone. The language you ship in.
Every lesson's sandbox runs in Python, TypeScript, or Java, so you practice in the stack you already work in, wherever you are.
Start at the bedrock. Build to the frontier.
The first part is free. Open the curriculum, run a real lab in your first five minutes, and start the climb to staff-level AI engineering.