About Me

I'm a graduate researcher in computer science at UniversitĂ© Laval, working at the intersection of machine learning and bioinformatics. I lead the ML side of a closed-loop antimicrobial peptide (AMP) discovery system that pairs model-guided candidate selection with an high-throughput wet-lab assay. I also study how data leakage and non-i.i.d. structure quietly break evaluation in biochemical ML — work that led to the QMAP benchmark and, more generally, the Refnd framework for leakage-safe splitting. Appart from BioAI, I am also interested in robotics. I am collaborate part-time with NORLAB on visual SLAM robust to high-dynamic-range scenes. Finally, I also lead a student team building autonomous FPV drones for a laser-tag competition.

My goal is to build ML systems that integrate tightly with wet-lab validation, creating a self-reinforcing loop where the model learns from its own mistakes. I believe this synergy between computation and experiment is key to accelerating therapeutic development. I'm interested in drug design broadly — peptide-based drugs, gene therapies, antibody therapies, and RNA therapies alike, with a particular pull toward gene therapies.

Highlights

  • Closed-loop ML/wet-lab AMP discovery system (ML lead) — ~25% anticipated hit rate.
  • 3x first-author scientific contributions (Scientific Reports, ASMS poster, technical report targeting ICLR).
  • NSERC + FRQNT funded (twice).
  • 4.33/4.33 (M.Sc.), 4.29/4.33 (B.Sc.), UniversitĂ© Laval.