I think deeply about how things work and how they could work better.

I'm currently researching agentic memory at The Matter Lab, under Prof. Alan Aspuru-Guzik. I lead the design and development of the memory infrastructure, evaluation harness, and benchmark suite for the agentic memory subsystem behind El Agente.

Since September 2026, I've also been researching controlled probing frameworks for deep neural networks, with an initial focus on transformer-based vision models. I'm interested in not only what these models learn, but how we can study their internal representations carefully and draw conclusions we can trust.

I started in pure math, spending hours studying theorem proofs and learning to think carefully about structure, logic, and first principles. Since learning to code, I've been drawn to engineering work that begins with a deep understanding of the problem. I believe the best systems and the most original ideas come from clear foundations.

I'm now pursuing an Honours BSc at the University of Toronto, studying computer science, mathematics, and statistics. I like moving between ideas and implementation: understanding the assumptions behind a model, building the system around it, and designing experiments that reveal where it can be improved.

Much of my favourite work is under NDA or lives in private repositories (papers coming soon 👀), but I hope you enjoy the smaller projects I've set aside time to share here! ^^