MSc student in Research and Innovation in Computer-Based Science and Engineering (RICSE) at Universitat Autònoma de Barcelona, and part-time research engineer at IEEC. I like making heavy computations fast: GPU programming, AI acceleration, numerical simulation and large data pipelines.
Looking for R&D roles in GPU computing, HPC and applied AI.
├─ r2net-cuda/CUDA acceleration of the R2-Net neural lossless image coder, comparing wavefront and fully parallel GPU strategies. Python · CUDA · CuPy · NumPy
|
├─ neural-fluid-simulator/A U-Net trained on lattice Boltzmann simulations that predicts vortex shedding 52× faster, on obstacles and Reynolds numbers it never saw. Python · PyTorch · NumPy
|
├─ kerr-black-hole-simulator/Real-time Kerr ray tracer in GLSL that follows light through curved spacetime to render gravitational lensing and the accretion disk. Validated against analytical results. C++17 · GLSL · raylib · CMake
|
╰─ survey-tabulator/Desktop tool that turns SPSS .sav surveys into weighted frequency tables (Excel and PDF). Tabulates two-million-row files in 15 seconds without loading them into memory.Python · pandas · pyreadstat · ReportLab
|
-O3 and a lot of curiosity · @ablesarodriguez