Louen Pottier
AI Research Engineer · Scientific ML, Physics & 3D Vision
Research engineer at the intersection of physics and deep learning. Graduate of ENS Paris-Saclay, I have been working for three years at CEA designing neural network architectures for real-time physical simulation in virtual reality. My work has led to several scientific publications and a patent filing. Along the way I have also worked on applying deep learning to a wide variety of problems: geolocation, 3D reconstruction, verification of industrial PLC programs or car crash simulation. I enjoy exploring new fields and finding unexpected connections with my past work.
Currently open to new opportunities, and always up for discussing a scientific problem. Feel free to reach out!
Highlights
XR DAYS 2026 · La Ruche Industrielle, Vénissieux
LaGSplat live demo at XR DAYS 2026
A 3D scene made physical from videos alone with LaGSplat: immersed in XR, a user can interact with the scene, whose physics LaGSplat simulates in real time.
Preprint arXiv:2608.16324 · patent filed
LaGSplat: Latent Lagrangian Gaussian Splatting
Method to build a physics-governed interactive simulation from monocular video. Makes it possible to apply forces inside a video without a single force ever being measured.
Mistral Vibe Hackathon (2026) · 3rd overall
Impostral
Multiplayer Turing test: Mistral agents infiltrate a group of humans and try to blend in. Can you unmask the AIs before they convince everyone you are one? 🏆 1st in our category and 🥉 3rd overall, winning €1,000 in Mistral credits.
Experience
CEA-LIST
Research Engineer
Design of neural network architectures for physical simulation and model reduction applied to virtual reality (physics-informed NNs, surrogate models, world models, Gaussian Splatting). One patent filed and several scientific publications. Industrial consulting: finite element simulation substitution for Renault (crash test, GNN) and Forvia (AI for design optimization).
Read my publications → Browse my projects →
ESILV
Part-time Lecturer
Teaching fluid mechanics, thermodynamics, and partial differential equations to engineering students at Bachelor and MSc level. Courses delivered in both French and English.
See my teaching →
EDF Lab
Research Intern
AI for fluid mechanics at EDF Lab Paris-Saclay: continuous-kernel convolution layers for irregular meshes, applied to the prediction of vortical flows. Results presented at the NeurIPS 2021 workshop on Machine Learning and the Physical Sciences.
Read the paper →Skills
Machine Learning & SciML
PINNs · Surrogate Models · Representation Learning · 3D/4D Gaussian Splatting
Model Deployment
PyTorch · C++ · ONNX · TorchScript · Libtorch · Docker
Physics & Simulation
Continuum Mechanics · Model Order Reduction · FEniCS · Cast3M · Ansys
Languages
French (native) · English C1 - Cambridge English Qualification
Education
CentraleSupélec
M.Sc. - Mechanical Simulation of Structures & Coupled Systems
Specialisation in computational mechanics, finite element methods, and coupled multiphysics systems, with coursework in AI for physical simulation.
ENS Paris-Saclay
Diploma in Artificial Intelligence
Selective internal programme at ENS Paris-Saclay. Second year: research internship at EDF R&D, co-supervised by Centre Borelli (ENS AI research lab), with coursework from the MVA master's programme. Work presented at a NeurIPS workshop.
ENS Paris-Saclay
M.Sc. - Mechanics of Materials and Structures
Nonlinear mechanics, damage and fracture, continuum mechanics, and numerical methods. Simultaneously followed the first year of the ENS AI diploma (selective internal programme).
ENS Paris-Saclay
B.Sc. - Mathematics, Physics & Engineering Sciences
Intensive undergraduate curriculum in mathematics, physics, and engineering sciences at ENS Paris-Saclay (SAPHIRE programme).
Lycée Blaise-Pascal
Classe Préparatoire aux Grandes Écoles
Intensive two-year preparatory programme in mathematics, physics, and engineering sciences.