Projects
About Me
As a kid I grew up in rural Canada without internet, so I passed the time building stuff out of scrap. Including motorcycles, rocket-powered GT racers, and drift trikes. I also have a 1986 MR2 I've been building since I was 16. Once I moved to Saskatchewan for my Bachelor's degree and finally got internet access, I could finally delve into programming. My focus was on artificial intelligence, specifically genetic algorithms and neuro-symbolic agents.
I'm a big fan of unconventional thinking and solutions. My past projects have included a stock trading bot and a basic language model trained to mimic TV characters. During my exchange at the University of Tsukuba, I designed a safety-oriented A* pathfinding system using taxi crash data from Tokyo to help recommend safer driving routes. Later on, I worked on a dynamic MNIST classifier that could prune and regrow neural connections based on how useful each node was, mimicking a kind of artificial apoptosis.
I graduated from the University of Saskatchewan in 2025. Following this I created a deep Q-learning agent in a Minecraft environment to study open world exploration. I'm especially interested in how symbolic logic can be passed from agent to agent through interactions and genetics.
Research
Right now, I'm working on combining neural networks with symbolic logic to make agents that can reason and adapt much like animals and humans can. To test them, I've been developing a 2D sandbox environment using Pygame, kind of like a stripped down Terraria. The idea is to use this space as a playground for AI experiments as I head into my Master's research.
Blog
Photos
