I am an AI researcher interested in understanding higher-order cognitive functions and in developing learning systems that reflect the structural principles underlying human intelligence.
I am currently a Master’s student at Mila and Université de Montréal, supervised by Guillaume Lajoie and Dhanya Sridhar. My research explores compression as an important organizing objective in intelligent systems, and how it shapes representation, abstraction, and generalization. In particular, I study how compression-driven learning objectives can induce compositional structure, encourage the reuse of learned components, and give rise to algorithmic behavior in models trained on structured data. I am especially interested in how data organization and learning curricula influence what models learn and how efficiently they can generalize.
Previously, I interned as a Machine Learning Engineer at Intellisports Inc., where I worked on human activity recognition using sequence models, and at MiQ, where I developed scalable machine learning pipelines in large-scale data environments.
Outside of research, I enjoy chess, climbing, snowboarding, and playing football.
