I'm a 15-year-old from India teaching myself machine learning from the ground up. My method is simple: learn a concept, implement it in a project — and let that project reveal the next thing worth learning.
I build alongside AI and learn from what ships. Projects aren't a distraction from the study plan — they are the plan.
When I'm not building, I'm studying or on court.
Never get stuck in the course loop — at some point you have to build.
A reinforcement-learning agent that taught itself a number-guessing game — plus the playable web game, deployed to the world. Trained the model, built the frontend, shipped it. End to end.
A cooling-optimization evaluator for data centers — the infrastructure that every model I train has to run on. Scoring thermal strategies before they ever touch real hardware.
Not just an engine that plays — one that explains. Move-by-move analysis in plain English, built as the stepping stone toward the dream project: a self-play RL chess engine.
Consistency beats intensity — show up daily.
Footwork, patience, and long rallies — the physical counterweight to hours of building.
🎾Strategy over 64 squares. It's no coincidence the dream project is an RL chess engine.
♔