Cosmology from galaxy clusters and the CMB.

I'm based at the Cavendish Laboratory, University of Cambridge, where I work on cosmology, mostly pulling cosmological information out of galaxy clusters and the cosmic microwave background. I'm also affiliated with the Kavli Institute for Cosmology and the Infosys-Cambridge AI Centre.

At the moment I'm working on confirming clusters detected via the thermal Sunyaev-Zel'dovich (tSZ) effect, using machine learning. tSZ surveys throw up a lot of candidates, but not all of them turn out to be real clusters, so getting that confirmation step right matters if the resulting catalogue is going to be much use for cosmology. I'm supervised by Boris Bolliet.

Alongside that, I work on agentic AI for scientific discovery in cosmology: systems like CMBAgent that can plan an analysis, write the code, run it, and interpret the results, more like a research collaborator than a script you run once. I'm interested in how far that kind of agent-driven approach can go, and what it means for the way cosmology gets done.

This all started with my MPhil dissertation in Data Intensive Science at Cambridge, where I looked at parameter inference using diffusion models. It's stuck with me since, in how I think about using machine learning for cosmology.

Outside research, I enjoy sport and can often be found rowing on the River Cam. I also love being outdoors, whether that's hiking a new trail or exploring a city I haven't visited before, so my weekends and holidays tend to be spent outside wherever possible.