Bio
I am a Research Associate at The Alan Turing Institute in London, in the AI for Physics team of the Fundamental Research mission, where I work with Jason McEwen. I completed my PhD in computational cosmology in October 2026 at the AstroParticule et Cosmologie laboratory (APC, CNRS/IN2P3) and Université Paris Cité, advised by Eric Aubourg, Alexandre Boucaud, Josquin Errard, and François Lanusse. My thesis, “Differentiable Field-Level Cosmology: CMB, Weak Lensing, and GPU-Accelerated Inference”, develops the differentiable simulation and inference methods needed to extract cosmological information from the next generation of CMB and large-scale-structure surveys.
Interests
My main interest is the intersection of physical models and machine-learned emulators for weather forecasting on spherical geometries. I am also interested in Bayesian uncertainty quantification for high-dimensional problems. Such problems require differentiable forward models and scalable samplers, which I built in JAX for cosmological inference during my PhD.
Research
At the Turing, my goal is to build emulators for weather forecasting on spherical geometries that remain consistent with the underlying physical models, and to quantify their uncertainty with Bayesian methods.
During my PhD, I developed differentiable N-body solvers, most notably JaxPM, which I maintain, at the scale of Rubin LSST. Multi-node domain decomposition and automatic differentiation allow hierarchical Bayesian inference directly over forward simulations of large-scale structure, which makes full-field inference and near-optimal extraction of cosmological information possible.
As a member of the Simons Observatory, I also worked on CMB component separation, focusing on the spatial variability of foreground spectral indices, which biases the tensor-to-scalar ratio if left unmodelled. I developed a fast, scalable minimiser in the FURAX framework that remains robust in the extremely noisy regimes expected for upcoming full-sky missions such as LiteBIRD.
Prior to academia
Before starting the PhD I spent five years in industry. I was a Software Infrastructure Engineer at Dassault Systèmes, where I worked inside the multi-million-line CATIA C++/Linux codebase, and earlier a Data Acquisition Engineer at Renault. Those years shaped how I write research code today.