- Developed deep Bayesian neural network architectures, including dynamics modeling, probabilistic classification, and attention mechanisms.
- Implemented core CUDA kernels for the cuTAGI open‑source library (Tractable Approximate Gaussian Inference).
Work
- Mar2025 - CurrentPolytechnique MontréalPhD in Computer Engineering
- Jan2024 - 04/2025Polytechnique MontréalM.A.Sc. in Computer Engineering
- Derived mathematical formulations to quantify uncertainty in diffusion models.
- Contributed to C++/CUDA/Python development of the cuTAGI library, enabling uncertainty capture in diffusion models and deep neural networks.
- Jan2019 - 06/2023Universitat Politècnica de CatalunyaB.S. in Computer Engineering
- Specialized in Computer Science.