I've just finished an MSc in Artificial Intelligence and Machine Learning at Queen Mary University of London, after a First-Class BSc in Computer Science at the University of Portsmouth. I build machine learning systems end to end, from data pipelines to trained models and deployment.
My work spans deep learning, reinforcement learning, time series and physics-informed neural networks, applied to medical imaging, finance and space science.
I'm looking for ML and AI engineering, data science and quant research roles in London.
- RL vs Traditional Portfolio Allocation: my MSc dissertation. PPO agents allocating capital across 119 assets, tested against Markowitz, minimum variance, risk parity and 1/N with point-in-time data and realistic trading costs.
- Pneumonia Detection CNN: my BSc dissertation. A CNN built from scratch in NumPy and Numba, with no deep learning frameworks, classifying chest X-rays.
- Cloud Fraud Detector: a PySpark fraud classifier on GCP, provisioned with Terraform, containerised with Docker and scoring transactions in real time from Pub/Sub. AUC-ROC 0.96.
- AI in Astrophysics and Space Science: eight MSc assignments applying Bayesian models, Gaussian processes, CNNs and physics-informed neural networks to space physics data.
- URA Volatility Forecasting: a stacked LSTM forecasting 21-day realised volatility on a uranium ETF, benchmarked against GARCH(1,1).
- Market Making Hackathon: price prediction and quote generation for live market-making rounds.
- Languages: Python, C++, SQL, JavaScript
- ML: PyTorch, TensorFlow, scikit-learn, NumPy, pandas
- Data and cloud: PySpark, GCP, Docker, Terraform, Git
- Methods: deep reinforcement learning, CNNs, LSTMs, physics-informed neural networks, time series (ARIMA, GARCH), Bayesian inference