Suggested Paths

For beginners, the biggest risk isn't picking the "wrong" topic — it's overwhelm and quitting. Pick a goal, follow the order. Each path is complete on its own — it includes whatever it needs.

Pick a path

AI Engineer (product)

~8,593 lessons · about 7 months at 10 hrs/week

Different from the Data/ML track — skips the mathy ML internals and focuses on integrating models into products. What most AI engineering jobs actually are right now.

Log in to save this path to your profile and track progress.

  1. 1

    Python

    Programming Languages · Beginner · 1,060 lessons

    The main language of AI tooling — SDKs, evals, and data scripts are mostly Python.

    Start course →

  2. 2

    Git & GitHub

    Software Engineering · Beginner · 96 lessons

    Short and practical: commits, branches, and GitHub — save your work properly from day one.

    Start course →

  3. 3

    Software Engineering

    Software Engineering · Intermediate · 1,933 lessons

    Testing, code review and CI — AI features need the same engineering discipline as any other code.

    Start course →

  4. 4

    Databases & SQL

    Databases · Intermediate · 1,953 lessonsthrough: 6. Query Performance and Indexing

    App data and vector search both live in databases. Schemas, queries, and indexes.

    Start course →

  5. 5

    AI Engineering

    AI & Data · Advanced · 1,886 lessons

    The main event: LLMs, RAG, agents, evals, production AI patterns.

    Start course →

  6. 6

    Machine Learning and Deep Learning

    AI & Data · Advanced · 3,538 lessonsthrough: 2. Machine Learning Mindset and Core Concepts

    A light dip for context — enough to know what's happening inside the models you're calling. Don't go deep.

    Start course →

  7. 7

    JavaScript

    Programming Languages · Beginner · 2,723 lessons

    Optional, once Python feels solid: most AI products ship as web apps, so learn the language of the browser next.

    Start course →

Beginner-protection rules

Apply these regardless of which track you pick.

  • Pick one path and commit for at least a month. Overwhelm and quitting are the real risks — not picking the 'wrong' path.
  • Learn one language at a time. Switching languages too early kills momentum.
  • Build something small after every course — reading alone won't make it stick.
  • Short, steady sessions beat weekend marathons. A few lessons a day adds up fast.
  • You don't need to finish every lesson before moving on — go deep where your goal needs it, skim the rest, and come back later.
  • Paths that include Mathematics or Computer Science use them as tools. Learn alongside your main courses; don't let them block you.
  • Ship one real project before diving deep into System Design or Security — they make far more sense once you've built something.

Want to skip around instead? Browse all courses.

New to programming?

Optional · about 3 months at 10 hrs/week

Optional. Never written code before? Python gives quick wins and CS gives you the vocabulary programmers use. Already code? Skip this and jump straight into a path. If your path starts with JavaScript (Web, Full Stack, Mobile), skip Python here — one language at a time.

  1. 1

    Python

    Programming Languages · Beginner · 1,060 lessons

    First language. Quick wins build the habit before anything harder.

    Start course →

  2. 2

    Computer Science

    Foundations · Intermediate · 2,872 lessons

    Light pass — variables, functions, data structures, complexity. The vocabulary downstream courses assume.

    Start course →