Summary
Jean-francis Roy is an R&D Director and Staff Applied Scientist with 16 years of experience building and leading machine learning teams at Coveo, combining deep academic rigor from a Ph.D. in computer science with practical production delivery. His doctoral work introduced a family of PAC-Bayesian–justified ensemble algorithms that optimize voter disagreement and remain competitive with state-of-the-art methods, reflecting a strong foundation in theoretical ML and reproducible research. At Coveo he progressed from research intern to director, shipping ML-driven search features and scaling research into product impact while mentoring teams. He also has a history of teaching core CS courses and contributing to open-source Linux projects, demonstrating both pedagogical strength and systems-level curiosity. Based in Quebec City, he blends research, engineering, and team leadership to turn advanced learning algorithms into usable, business-facing systems. A less obvious trait: his background spans low-level systems (Funtoo/GNU Linux) to high-level statistical learning, giving him rare fluency across the full ML stack.
16 years of coding experience
6 years of employment as a software developer
B.Sc.A, Computer Sciences, B.Sc.A, Computer Sciences at Université Laval
DEC, Informatique, DEC, Informatique at Collège François-Xavier-Garneau
Machine Learning, Machine Learning at Machine Learning Summer School
French, English