3rd-year Computer Science at Queen's University
Toronto, ON
I'm currently a part-time researcher at Université du Québec, studying the security of LLM-generated code. I like working with the newest tech, not to jump on the hype train, but to see what it can actually do out in the real world.
I spend most of my time writing software, but when that starts to feel too repetitive, I turn to hardware and try to mix the two. I don't like sitting still. I'd rather be building, learning, or chasing the next idea that makes me think, "What if this could actually work?"
Part-time Researcher
Security research on LLM-generated code. Second author of WildCode (FPS25), using CodeQL and OpenGrep for static analysis.
2025 to Present
Junior AI Policy Analyst
Built a local RAG dashboard and an interactive data-entry platform with sovereign LLMs. Assessed LLM safety and cyber model capabilities for policy decisions.
Jun 2026 to Aug 2026
Researcher
Built BLADE, an LLM-free bug localization pipeline with 84% Top-10 accuracy across 463 bugs and 72% automated repair accuracy.
May 2025 to Aug 2025
Used by 4,900 Queen's students (about 1 in 4 undergraduates). Builds conflict-free timetables, checks prerequisites and corequisites, and accounts for travel time between main and West Campus. Built to replace slow, manual course planning, and featured in The Queen's Journal.
An end-to-end TinyML system for real-time autonomous navigation on an ESP32, running a lightweight CNN within 1MB of RAM. Includes a custom-built dataset and on-device inference, with the work published on arXiv.
1lb combat robot designed, built, and tested with 3D printing and CNC milling. Competed at Bingemans, went five fights before catching on fire, and won the "Best Death" award. A hands-on mix of mechanical design, electronics, and rapid iteration.