13.8 Kaggle and open source
You have public evidence of working with others' code.
Public evidence of working with others' code — a real leaderboard position, a merged pull request to an ML library, something small you maintain — is what separates learned-alone from can-work-on-a-team in a reviewer's eyes. It closes the course by opening the door outward. The waiting trap is feeling not ready: documentation fixes and good bug reports are legitimate first contributions, and the readiness arrives from contributing, not before it.
Work through these
Working a competition to a real leaderboard position
Working a competition through to a real position rather than a submission, which teaches evaluation and iteration under a hard measure. The position matters less than having finished.
Reading and learning from public notebooks
Reading published solutions is one of the fastest ways to learn practical technique. It is also where most of the competition's value actually sits.
First pull request to an ML library
A first contribution to an open library, however small, which teaches reading unfamiliar code and working to somebody else's standards. Both are what employment mostly consists of.
Maintaining something small of your own
Maintaining something of your own, even a small tool, shows a different capability from building one. Responding to somebody else's issue is the part that counts.
Sign in to keep your progress.
Free resources
We haven't checked most of these for screen reader use yet.
Links last checked 29 Aug 2026.
Stuck here?
Ask a mentor. A real person answers, and they can see exactly which topic you're on. Usually within a couple of working days.
Checking your session…
Topics shown in module order.