Raul Pelaez
Research Scientist at Mistral
Paris, Ile-de-France
Summary
👤
Senior🎓
Top SchoolRaul Pelaez is a research scientist and GPU software engineer with 11 years of experience accelerating compute‑intensive workloads across drug discovery, fluid dynamics, and ML. He designs and ships CUDA/C++ libraries from scratch (e.g., UAMMD), builds reproducible CI/CD and conda‑forge packaging, and has driven 50× speedups in real molecular‑modelling pipelines like TorchMD‑Net. Comfortable moving from hand‑optimised kernels to TypeScript front ends and WebGPU/WebAssembly apps, he also operates and benchmarks on local GPU clusters and integrates LLM tooling for research workflows. A community-minded academic who organizes conferences, mentors PhD students, and lectures at IE University, he brings production-grade engineering practices to research code. Notably, he combines low‑level CUDA graphs and Autograd‑enabled ops with cross‑platform packaging to make GPU research reproducible and widely deployable. Based in Madrid, he seeks roles that push GPU hardware limits in both exascale and cloud environments.
11 years of coding experience
12 years of employment as a software developer
Doctor of Philosophy - PhD Computational physics, Doctor of Philosophy - PhD Computational physics at Universidad Autónoma de Madrid
English, Spanish
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (7)
conda-forge10
bash10
build-automation10
cmake9
python8
dockers7
docker7
Programming languages (16)
Github contributions (5)
conda-forge/staged-recipes
Sep 2023 - Sep 2025
A place to submit conda recipes before they become fully fledged conda-forge feedstocks
Role in this project:
DevOps Engineer & Build Engineer
DevOps Engineer & Build EngineerContributions:5 reviews, 4 PRs, 32 comments in 1 year 11 months
Contributions summary:Raul contributed to building and configuring conda recipes within the conda-forge ecosystem. The contributions focused on modifying build scripts, specifically for `torchmd-net` and `openmm-xtb`, including adding tests, and enabling the build process for different configurations. These changes involved setting environment variables for CUDA architectures, integrating Python packages, and configuring build processes through CMake.
condaconda-forge
stochasticHydroTools/libMobility
Dec 2021 - Apr 2026
A unified interface to compute hydrodynamic displacements.
Contributions:17 releases, 48 reviews, 48 PRs in 4 years 4 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.