Every module is a visualisation you can drive, running the real computation — machine learning, deep learning, algorithms, and the maths underneath them. No jargon, no account, nothing gated.
Start here
Each of these is the opening module of its track - the one that assumes the least.
Try adjusting your search terms or checking another category.
Every track
Distance from the centre is the number of modules in that track. Point at a spoke, or tab through them, to pull one out - each is a link to that track.
Hover a spoke for its count
| Track | Modules |
|---|---|
| Gen AI | 52 |
| Interview | 49 |
| Computer Vision | 47 |
| Maths | 47 |
| Deep Learning | 46 |
| Python | 46 |
| Machine Learning | 45 |
| Database | 42 |
| Algorithms | 41 |
| NLP | 41 |
| Pydantic | 30 |
| FastAPI | 26 |
| pandas | 26 |
| scikit-learn | 23 |
| matplotlib | 22 |
| NumPy | 20 |
| Async Python | 5 |
How it works
No account, no install, nothing gated. Open a page and it runs.
Every module is one concept with one visualisation. There is no chapter to finish before it makes sense, and no order you have to arrive in.
The controls are the point. Set k to 1 and watch KNN overfit, push the learning rate until descent diverges. Understanding usually arrives at the failure, not the happy path.
Where it helps, the same page carries a Python editor running the real thing in your browser. Change a line, run it, see the output move.
Questions
Yes, and there is nothing to sign up for. No account, no trial, no email wall.
It runs CPython compiled to WebAssembly, in a worker thread in your browser. Nothing is uploaded and nothing is executed on a server.
The site installs as an app and caches what you have opened, so pages you have already visited keep working without a connection.
Anyone who has read the definition and still cannot picture the thing. It assumes no maths degree and leaves nothing as an exercise for the reader.
That is the whole pitch. 608 modules, all of them running the real computation, none of them asking you for anything first.