Learn data skills by building real projects
Why this is different
Most “learn data” courses have you watch someone else type. Here, you build. Every project hands you a realistic dataset and a decision a working analyst, engineer, or data scientist actually has to make — then walks you to a correct, tested answer. The skill you leave with is judgment, not syntax you’ll forget.
- Run it in your browser. On projects that support it, click ▶ Run and real Python executes on the page — nothing to install.
- Or run it locally, with copy-paste setup for Windows, macOS and Linux.
- No account, no
git, no GPU. The code comes to you from this site. - Every number is real. Runtimes, row counts and results all come from an actual run.
- Realistic datasets with a known answer. Mostly purpose-built (generated, never scraped, no personal data); a few use a classic, cited public dataset. Either way the correct answer is known, so you can prove your method actually works — and the skill you build is exactly the same as on real data. How this is made.
- Two parts each: a guided Part A anyone can finish, and an optional Part B to stretch.
- Free forever. Free tools only — no trials, no credit card, no bait-and-switch.
How it works
Pick a project
Choose a track and a difficulty. Each project page tells you the skill, the time, and what you’ll build before you commit.
Build Part A
Follow the guided tutorial. Where a project has a ▶ Run button, click it to run in your browser; otherwise run it locally — checkpoints tell you you’re on track.
Stretch with Part B
An optional, tougher variant with a solution you can check yourself. This is where the skill sticks.
Six tracks, beginner to advanced
Data Analytics
Find what a dataset is really telling you — and where it quietly lies.
Data Engineering
Move and reshape data reliably: contracts, schema drift, clean loads.
AI & Machine Learning
Train models and — more importantly — learn when they’re fooling you.
Supply Chain
Inventory, routing and stocking decisions modelled on real trade-offs.
Geospatial
Work with places: joins, clustering, and distance done correctly.
Open-Source Libraries
Get fluent in the tools pros reach for: DuckDB, Polars, networkx, more.
Who it’s for
You’re comfortable-ish with a computer and curious about data. Maybe you’ve done a bit of Python, maybe not. You don’t want to sit through more theory — you want to make something and understand why it works. Start with the on-ramp and climb from there; every project names its prerequisites so you’re never lost.
Ready to build something?
Start with a two-minute orientation, then run your first real project — in your browser or with a quick free setup, in the next few minutes.
Start here →