cv

Senior ML/Research Software Engineer

Institute of Computing for Climate Science (ICCS)

Specialised in the application of machine learning and high-performance computing techniques to solve complex modelling problems in climate science.

• Led a team of 3 engineers to design and deliver a modular forecasting and physical evaluation framework of AI-based emulators for ocean spin-up workflows.
• Coupled PyTorch models with legacy Fortran climate simulation codes using FTorch.
• Supervised and mentored junior RSEs on distributed training best practices for ML engineering in scientific computing

Computational Scientist / Research Software Engineer

University Information Services (UIS), University of Cambridge

Led development of research computing solutions spanning CFD, FEA, and ML workloads on HPC platforms.

• Built and evaluated large-scale ML pipelines across HPC systems, including web-scale NLP workflows using Common Crawl data (multi-terabyte scale). Explored and compared GraphSAGE, Doc2Vec, and TF-IDF embeddings for downstream document and graph classification tasks.
• Worked on the performance optimisation and scaling of deep learning models (e.g., AlphaFold2, vision, and LLM inference) across heterogeneous hardware platforms and interconnects.
• Developed a hybrid FEM-BEM solver for wave propagation using FEniCS and BEMpp. Designed a parallel architecture using MPI communicator splitting (32:1) to balance load between coupled solvers.
• Set up CI pipelines and reproducible ML/HPC benchmarking infrastructure FEniCS benchmarking using GitLab CI and reframe, to support model and solver development on pre-exascale supercomputers.

PhD

University of Cambridge

Physics

MPhil

University of Cambridge

Scientific Computing

Grade: Distinction

MSc

University of Manchester

Computer Science

Grade: 1st

MEng

University of Sheffield

Chemical Engineering and Chemistry

Grade: 1st

Tree crown detection using Mask R-CNN

Remote Sensing in Ecology and Conservation

Contributor

Pushing the limits of exoplanet discovery via direct imaging with deep learning

ECML PKDD, Springer

Contributor