Probably the most un-physicist physics student you'll ever meet. I don't sit around deriving Lagrangians for fun. I build things. Trading bots that actually make money, AI tools that reconstruct faces from witness descriptions, pricing engines that run in your browser. If there's math behind it, I've probably tried to code it.
The thread that connects everything I do is taking stupidly complex problems like stochastic calculus, neural network optimisation, and financial derivatives, then turning them into something that actually works and ships. I've deployed models that trade live markets, built hackathon projects that won prizes, and I'm currently finishing my MSci at Imperial while targeting publications in ML and computational finance.
I'm not neatly one thing. Quant researcher, full-stack developer, ML engineer, physics student. I'm all of them, and none of them fully describe what I do. Currently obsessed with: rough path theory in neural networks, LLM hallucination benchmarks for finance, and making Rust do things it probably shouldn't.
Predicted 1:1 (First Class Honours)
A*A*A* in Mathematics, Physics, Computer Science
13 GCSEs with grades 9 to 8 including Maths and English, and 2 Distinctions
From particle physics simulations to quantitative finance models, explore my diverse portfolio of technical projects
AI-powered forensic tool transforming witness descriptions into photorealistic suspect images, mugshots, video, and 3D models
On-chain PvP price prediction duels with trustless settlement via Flare's enshrined oracle protocols
Interactive derivatives pricing engine with 15+ SDE models, built in Rust/WASM with a Bloomberg-inspired UI
Open-source SEC finance Q&A benchmark for evaluating and mitigating LLM hallucinations
First production-ready algorithmic trading system harnessing social media virality for financial alpha
Low-power 8-bit microcontroller running real-time ML for autonomous gameplay
Advanced options pricing integrating stochastic volatility and ML predictions
Extracted neutrino parameters θ23 and Δm²23 using advanced optimization algorithms
Real-time market predictions combining traditional data with sentiment analysis
Modern portfolio featuring 168 floating physics equations, 5 hidden easter eggs, and a black hole
ML-powered detection of staleness and false jumps in fixed-income data feeds
Simulated discovery of Higgs boson using advanced statistical methods and Python
C++ risk management system calculating VaR/CVaR with ML predictions
Hard-sphere collision simulation validating classical gas laws
Mechanical stabilizers using gyroscopic principles for precision instruments
ML-powered rent calculator using hedonic regression for mathematically fair pricing
Interactive physics education tool with adaptive AI difficulty
Automated tool to download Spotify playlists via YouTube with metadata
Exploring the intersection of physics, mathematics, and computational finance through rigorous research and innovative solutions
EMNLP 2026 (Target) • 2026
Manually curated and open-sourced FinHALO-500, a 500-pair SEC finance Q&A benchmark and mitigation library, introducing two novel evaluation metrics: NIWH (materiality-weighted errors) and BAHE (cost-performance optimisation). Developed end-to-end mitigation pipeline combining RAG with QLoRA fine-tuning, reducing GPT-4o hallucinations by >20pp on SEC filing numeric tasks while slashing inference costs by 78%.
Journal of Financial Data Science (Target) • 2026
Novel approach to volatility forecasting combining LSTM networks with traditional GARCH models. Emphasis on high-frequency trading data and market microstructure.
JMLR / NeurIPS 2026 (Target) • 2026
Theoretical framework connecting path integrals from physics to neural network training dynamics. Applications to understanding loss landscape geometry and optimisation trajectories.
Interested in collaboration or discussing research opportunities?
Get in TouchWhether you're interested in research collaboration, have a challenging problem to solve, or just want to say hi – I'd love to hear from you
Open to opportunities in quantitative research, data science, and software engineering roles.