Software, tools and projects I designed or led: scientific computing, visualization, endurance sport and health.
Scientific library · 2026
Senior Lead Developer and Architect
Fast two-dimensional discrete cosine transforms (DCTs) in C11 with SIMD acceleration, a NumPy interface and numerical verification tools.
Orthonormal DCT-II, inverse and round trip · 8 to 1024 points per axis, rectangles included · float32
C11 · SIMD · Python/NumPy · macOS, Ubuntu, Windows · BSD-3-Clause license
GitHub ↗Paper ↗Releases ↗60 views in 14 days
Web tools · Since October 2026
Initiator and project lead
Athlensys aims to help people manage their sports training better and take it further. It offers free, simple and well-integrated tools that bring together software engineering, perceptual interfaces and AI agents. First tools: predicting a finish time, planning a race and measuring one's margin with Réserve⁵. It is an ongoing project, with more tools and products to come.
TypeScript · React · in-browser computation · triathlon, swim, bike, run
Perceptual interface · 2026
Design and development
Réserve⁵ explores what I call a perceptual interface, applied to triathlon race planning: every choice of intensity has a consequence you perceive intuitively. The closer you get to your maximal effort, that is, your minimum time, the more agitated the membrane becomes; shapes and colours shift with the effort committed. The interface is also designed to be practical and intuitive. It was inspired by my own triathlon preparation and that of teammates.
Five views: Membrane, Pressure, Isochrone, Ring, Constellation · WebGL or SVG rendering · seven languages
TypeScript · React · WebGL · SVG · in-browser computation
Entrepreneurial project · 2022–2023
Co-founder and lead
Making services easier to access for people living with chronic low back pain. The project grew out of Concordia University's Life Sciences Entrepreneurship Development Program.
Team of three · user interviews · target market definition · first viable product · finalist at Forces AVENIR (Université de Montréal, 2022) and the McGill Dobson Cup (2023)
Ph.D., DIRO, Université de Montréal · since 2024
A tile-based method built on partitioning and joining, matching the state of the art in accuracy–runtime trade-off in the evaluated configurations.
View project →
Master's, Université de Montréal · 2012–2014
From Kinect depth data to perceptual maps, to quantify gait symmetry and detect pathologies.
View project →Polytechnique Montréal · 2020
Including super-resolution for quantitative susceptibility mapping.
View project →Research internship, CEA-Saclay · 2012
Detecting light sources from specular reflections on shiny objects, for photorealistic augmented rendering.
View project →