Thomas Bouvier

I collaborate with scientists to research, create, and optimize scalable, ML-accelerated pipelines—ranging from Graph Neural Networks to Large Language Models to Diffusion Models—that help answer broad scientific questions. I hold a PhD from Inria in High-Performance Computing (HPC) and am specialized in scaling AI workloads from one GPU to hundreds. In tandem with these efforts, I also work on low-level performance characterization on GPU platforms.

I am also interested in continual learning, big data analytics, large-scale pre-training, and reproducibility in computational sciences.

I am passionate about finding solutions to the climate crisis we face, both in policy and through the development of software tools.