Kevin Lessard

Kevin holds a master's degree in machine learning from MILA (University of Montreal), which he completed after earning his bachelor's degree in mathematics and computer science. Fascinated by algorithms, he shared his knowledge as a teaching assistant and developed a natural affinity for artificial intelligence, driven by the search for elegant and efficient solutions.

Throughout his career, Kevin has contributed to a variety of projects, from robotics to AI applied to civic issues. In robotics, his implementation of a simultaneous localization and mapping (SLAM) algorithm on Duckietown educational robots was selected as one of the platform's flagship projects. He has also explored the intersection of AI and democracy by developing a generative flow network (GFlowNets) approach to combat gerrymandering. His research in natural language processing allowed him to stand out during his internships before joining a startup specializing in neural rendering, where he develops 3D/4D reconstruction methods from real-world sensor data for validating physical AI systems. Despite the diversity of these projects, a common thread runs through them: applying technical rigour to problems that matter. This naturally led him to Baseline's mission and its collaborative approach to AI.

Outside of work, Kevin enjoys staying active through swimming, running, and cycling. A big fan of video games, he also dedicates time to reading and music—bass and piano—which has led him to compose with his band, with whom he occasionally plays in Montreal bars… if you happen to run into him!