CV

Employment

Scientific Software Engineer — Science & Technology Facilities Council (2024–Present)

  • Built machine learning pipelines and computer vision tools for multimodal nanoscale bioimaging within a cross-disciplinary team. Led the full project lifecycle: developing statistical models, implementing ML algorithms that scale to millions of measurements, and deploying automated pipelines that run reliably in production to turn raw, messy data into actionable scientific insights.

  • Built a production ML pipeline that achieved world-first 5 nm resolution in single-molecule microscopy by resolving systematic biases in conventional techniques. Developed in Python and MATLAB using maximum-likelihood estimation and Bayesian inference (nested sampling) for parameter estimation, providing rigorous uncertainty estimates for downstream clustering stages.

  • Developed a statistical analysis framework in Python (NumPy, SciPy, pandas, scikit-learn), including custom multivariate kernel density estimation with adaptive bandwidth control. Performed model selection by benchmarking competing physical models, and implemented a novel approach for handling correlated circular and linear statistics using insights from differential geometry.

  • Scaled ML workflows to HPC infrastructure, orchestrating thousands of parallelised simulations for model validation. Containerised using Apptainer (Singularity) to ensure reproducibility across computing environments. Managed compute resource allocation and job scheduling with Slurm. Designed and implemented a complete computer vision solution for automated multimodal image alignment.

  • Pioneered the use of micromachined fiducial markers to bridge optical and electron microscopy, enabling the previously-impossible tracking and extraction of regions-of-interest across modalities. Led the full development cycle from component design in OpenSCAD, coordinating production with the fabrication team, to detection and analysis in Python with openCV.

Postdoctoral Research Fellow — Institute of Physics, Academia Sinica (2022–2024)

  • Increased computational efficiency of numerical black hole simulations by a factor of 10, enabling the simulation of systems that would otherwise have been unfeasible. Deployed parallelised C code on HPC cluster, again managing compute resources and workflow orchestration with Slurm.

  • Led a team of 10 to produce a major catalogue of black hole simulations. Required strategic resource planning to efficiently utilise the supercomputing budget, and an adaptable leadership style across skill levels from junior researchers to senior experts.

Education

Ph.D. Astrophysics · Cardiff University

  • Aspects of the numerical simulation of binary black hole spacetimes

M.Sc. Theoretial Physics · King’s College London

  • Scalar Inflation and the Primordial Origins of Large-Scale Structure in the Universe

M.Sci. Physics · Royal Holloway University of London

  • Particle Collisions in the Vicinity of Black Holes

Skills

Python, Linux, Git, machine learning, scikit-learn, PyTorch, pandas, MATLAB, TeX, Bayesian inference, OpenCV, Bash, simulation, modelling, HPC, MCMC, nested sampling