Hello, I’m Fawz!
I’m a computer science student at McMaster University and a former software engineering intern at Fastlane in San Francisco.
I’ve worked across full-stack and forward-deployed engineering at several startups, building products from early ideas through implementation. My technical interests include multi-agent reinforcement learning, designing evaluations and interactive environments for agents, and learning under partial observability.
I do my best work on small, high-energy teams where people move quickly and take ownership. I’m especially drawn to early-stage startups with hard problems, direct communication, and little process for its own sake.
Outside of engineering, I play in the Collegiate Chess League and enjoy 20th-century history, poker, speedcubing, and weightlifting.
Projects
ML-Keeper: An interactive Unity benchmark for training and evaluating learned goalkeepers in football penalty shootouts. It pairs reproducible experiments with a playable demo.
Voyager: A cooperative survival environment for studying how decentralized RL agents coordinate under partial information. Agents must work together across procedurally generated islands to complete a shared rescue objective.
CFR-Poker-Bot: A local-first research toolkit for building and benchmarking heads-up no-limit Texas Hold’em agents. It supports the full loop from training strategies with counterfactual regret minimization to evaluating them in duplicate matches.
Helix: An execution engine that makes repeated LLM workflows faster by reusing work when inputs have not changed. It models tasks as a graph and reruns only the affected steps.