I’m a research associate at the Kavli Institute for Cosmology, University of Cambridge, and an AI researcher at PolyChord Ltd., working at the intersection of probabilistic machine learning and precision fundamental physics. My work focuses on scalable probabilistic inference to solve cutting-edge scientific inference problems. I completed my PhD at UCL while working on the ATLAS experiment at CERN, and have published work in top physics journals as well as presenting at top-tier AI conference venues (NeurIPS, ICLR). I’m part of Will Handley’s research group, where I build tools that make state-of-the-art uncertainty quantification practical for real-world scientific and engineering systems.

Current specific topics of interest:

• Cosmology, accelerated likelihoods for Gravitational Wave Physics and CMB physics.
• Simulation Based inference, interfacing generative AI with scientific inference problems.
• MCMC methods, particularly with a view to building scalable GPU inference tools.
• Particle Monte Carlo methods, Nested sampling and Sequential Monte Carlo

Current Work

I am currently building frameworks in JAX, a high-performance numerical computing library for machine learning research, and working extensively with BlackJAX, a library of samplers for JAX that provides state-of-the-art MCMC and variational inference algorithms.

I am happy to discuss any of these topics, and open to contact for collaboration.