I am a Ph.D. candidate in Physics at the Insitute for Computational Cosmology at Durham University, focusing on cosmology, large-scale structure and Bayesian inference. My research involves developing high-performance estimators for galaxy clustering and leveraging simulation-based inference techniques.

Before starting my Ph.D., I worked as a Quantitative Developer in investment banking (SMBC), specialising in xVA and derivatives pricing and as a Software Developer in financial technology. I hold an MSci in Physics from Imperial College London.

My academic research focus is on developing software tools for large-scale structure analysis and cosmological inference.

• SUGC – Sparse Unbiased Galaxy Clustering. High-performance, multi-threaded correlation-function counters in Rust and Python for calculating real-space, redshift-space, and high-order N-point statistics on galaxy catalogues..
• ANVIL – Amortised Noise Variance Inference Library. A high-performance framework for Noise-Amortised Neural Posterior Estimation (NPE) built with JAX, Equinox and Rust to amortise estimator variance in simulation-based inference.
• galform_execution & galform_analysis – Python utilities for configuring and executing the GALFORM semi-analytic galaxy formation model on HPC clusters, and analysing output HDF5 datasets.
• CMB_Advanced_Sampling – Cosmological parameter estimation from the Cosmic Microwave Background (CMB) using advanced sampling techniques.
• photon-bec-ml – Characterising the phase diagram of a photon Bose-Einstein condensate using machine learning classifiers.

Outside of physics research, I build high-performance desktop clients, automation tools, and financial infrastructure.

• aura – An automated assistant that daily fetches, filters, and summarises arXiv AI/ML publications using LLMs, emailing a structured digest.
• glide – A keyboard-first, multi-provider desktop email client for Linux built using C# with .NET 10 and Avalonia UI.
• synapse – A wealth management ecosystem combining portfolio tracking, ML-based equity forecasting, and mortgage tracking (TypeScript).
• heirloom – A fitness data aggregator that fetches, normalises, and deduplicates activity and health metrics from Strava and Garmin Connect into local SQLite storage.
• connect4-pic18-glcd – Connect 4 game written in PIC18F87K22 assembly language for a graphical LCD and a 4x4 matrix keypad.

• Durham University | Ph.D. in Physics (Cosmology & Bayesian Inference) | 2025 – Present
• SMBC | Quantitative Developer (xVA, derivatives pricing, trading tools) | 2024 – 2025
• Investment Software Ltd | Software Developer (financial software & pipeline optimisation) | 2022 – 2024
• Imperial College London | MSci in Physics | 2017 – 2021

• Languages: Python, Rust, C#, C, TypeScript, SQL, Assembly (PIC18), Bash, LaTeX
• Scientific Computing & ML: JAX / Equinox, PyTorch, TensorFlow Probability, NumPy, SciPy, AstroPy, Pandas, Scikit-learn
• Systems & Frontend: .NET 10, Avalonia UI, Substrate, Vue.js, Node.js
• Tools & Infrastructure: Git, Linux (Fedora), Docker, SQLite, PostgreSQL, SLURM (HPC)