Hi, I'm Sagar Pal

Computational Physicist • Scientific ML & HPC • Paris, France

I build high-performance computational systems at the intersection of physics and machine learning. Currently a technical lead at Entalpic, I help architect their ML platform for materials discovery — designing highly scalable pipelines that combine generative and predictive ML models with traditional HPC-based computational chemistry workloads.

My path here was unconventional. I started in mechanical engineering, moved into theoretical fluid dynamics, then spent my PhD at Sorbonne Université developing novel numerical methods for multiphase, multiscale, and multiphysics simulations — the kind of problems where air meets water at extreme speeds and everything fragments into droplets and sprays. During my post-doc, I pushed further into statistical physics, coupling direct numerical simulations with Monte Carlo techniques to model the stochastic nature of liquid fragmentation.

That journey gave me something rare: deep expertise in high-performance parallel computing, low-level systems and GPU programming, and numerical methods for complex physics — combined with the ability to translate all of it into modern scientific ML. I've worked across the entire stack, from hand-tuned Fortran running on 64,000 cores to PyTorch and JAX pipelines on GPU clusters. I care about building scientific software that's not just fast, but maintainable and adaptable to evolving research needs.

Download CV

Expertise

Domains

Scientific ML High Performance Computing Computational Physics Multiphase Fluid Dynamics Numerical Methods Statistical Physics ML for Materials Discovery

ML & Scientific Computing

PyTorch JAX Scikit-learn NumPy / SciPy

Systems & Parallel Computing

C++ Fortran Python MPI OpenMP CUDA