Three things have my attention: the air, the markets, and teaching. They're the same problem wearing different clothes.
I'm in Grade 9 at University of Toronto Schools. Everything I build starts with a claim I don't trust yet, and ends with me finding out whether it holds.
The air. In June 2023 Toronto's sky turned orange and the sensors that could have warned us had been reading the smoke for weeks. So I built CLEAR25, an early warning system that watches 440 stations and flags smoke before it lands. It's also personal. My brother Russell is in the CHILD Cohort Study at SickKids, which tracks children's lung health, so smoke isn't an abstraction in my house.
The markets. I've run a value-investing portfolio since I was ten, up 83% in five years, and I build backtesting systems to test whether a strategy survives real costs. Most don't. Finding that out cheaply is the whole point, and it's the same instinct that made me distrust my first smoke model.
Teaching. I tutor kids in English, math and science, prep students for the UTS entrance interview at Brain Power, and run HugoPrep. Before that I bought candy wholesale and sold it at school, which taught me more about margins than any book has.
Right now I'm training for the AI Olympiad and rebuilding CLEAR25's prediction layer as a graph neural network.