Tensor Golf — PyTorch tensor drills

Build fluency in PyTorch tensor operations through short, focused drills

Problem

Many learners understand ML concepts and math but get stuck the moment they need to write PyTorch tensor operations. There is no dedicated place to practice tensor manipulation; existing courses are theory-heavy, with few opportunities to actually write tensor code, so fluency never develops.

Solution

An online drill platform: each problem is a small tensor task (reshape, broadcast, index, reduce), solved by writing PyTorch code in the browser with instant validation; levels organized by topic and difficulty; accuracy and time tracked to build a daily practice habit.

Users

ML beginners, data science students, and developers transitioning into AI.

MVP

A built-in library of tensor-operation problems; in-browser code editor with instant judging; levels by topic; accuracy and timing records; a daily problem.

Revenue

Free basic problem set plus paid unlock of all problems and advanced topics, via subscription or one-time purchase.

Why now

The AI learning wave continues and PyTorch is the de facto standard, yet tensor-operation practice remains a gap. Both the problem bank and judging can run in the browser, so a solo developer can launch in weeks.