RESUME
Data science MPhil and Physics BSc with 2 years experience as a data scientist, including a range of machine learning projects. Looking to work on challenging projects while learning from experienced data scientists and machine learning engineers as part of a welcoming community.
Languages: Python (pandas, NumPy, Scikit-learn, Matplotlib, TensorFlow, PyTorch), Git, Docker, CI/CD, C, LaTeX
Education
- Modules: classical and Bayesian statistics, supervised & unsupervised machine learning, deep learning, high-performance computing.
- Thesis: Re-examining the putative radial velocity detection of L98-59b utilising a Gaussian Process framework: How reliably can we measure the mass of an exoplanet just half the mass of Venus? (85%)
- Modules: fluid dynamics in astrophysics, analysis and research for observational astronomy, mathematics, computational astrophysics, experimental physics & computing, quantum mechanics, thermodynamics, general relativity.
- Mathematics (A*), Physics (A), Art and Design (A)
- Silver award Physics Olympiad.
Experience
- Fine-tuned a LLM using AWS to perform classification tasks including sentiment analysis on video game chat forums.
- Constructed an image processing algorithm to filter thousands of Twitter images, compare to a leaked content database and automate takedown actions for Rockstar Games.
- Developed a "Digital Twin" model for retail logistics using gradient boosting and time series forecasting, deployed via Dash with a TensorFlow backend for real-time simulation.
- Researched and presented novel “Digital Twin” use cases, resulting in a project optimising chemotherapy patient journeys.
- Established a tutoring company that caters to 11+, GCSEs, A levels, STEP and Oxbridge preparation.
- Coached 36 students to date and organised a team of tutors: satoritutoringlondon.co.uk.
Projects
Machine Learning & Data Science
Academic Contribution
Contributed to Simon J. D. Prince's work on deep learning.
Cold Diffusion Models
Built a denoising diffusion probabilistic model (DDPM) from scratch, training a custom CNN to generate MNIST digits. Designed a cold diffusion variant with Gaussian blur and benchmarked performance using FID scores.
Air Quality Analysis
Applied unsupervised ML and time series methods to high-frequency air pollution data. Identified seasonal patterns, flagged anomalies, and proposed improvements to sensor calibration and spike detection.
Bayesian Inference for the Lighthouse Problem
Applied MCMC methods using the No-U-Turn Sampler to infer a lighthouse's location, based on a classic Cambridge problem by S. Gull and discussed in Data Analysis: A Bayesian Tutorial.
Astrophysics & Fluid Dynamics
Gaussian Processes for Exoplanets
Engineered a GP-based nested sampling system to isolate weak planetary signals from stellar noise. Applied quasi-periodic kernels to model stellar activity using HARPS/ESPRESSO observational data.
Exoplanet Analysis
Identified and investigated four "habitable" exoplanets using the Kepler dataset.
Runge-Kutta Method
Simulated Lagrange points in the Restricted Three Body Problem.
Monte Carlo Radiative Transfer
Simulated photon scattering through an atmosphere using Monte Carlo methods. Modelled isotropic and Rayleigh scattering to demonstrate the physical origin of sky colour and the role of mean free path in light diffusion.
Fluid Dynamics Simulation
Solved Euler's equations to simulate fluid dynamics in a shock tube.
Computing & Tools
High-Performance Computing
Leveraged GPU resources for astrophysics simulations and ML projects.
LaTeX
Produced technical reports using LaTeX.
Skills
Programming
- Python (pandas, NumPy, Scikit-learn, Matplotlib, TensorFlow)
- Git
- Docker
- CI/CD
- C
- LaTeX
Machine Learning
- Supervised & Unsupervised Learning
- Deep Learning
- Statistical Modeling
- Bayesian Statistics
Astrophysics & Physics
- Fluid Dynamics
- Computational Astrophysics
- Observational Astronomy
- Simulation & Modeling
Personal Interests
Played for a club growing up, as well as in school and university, and now play weekly with a team in South London.
Executed large-scale live music events, managing all aspects of preparation and leading teams to accommodate crowds of up to 600.