About Me

At University College London (UCL), my PhD research focuses complex problems in drug discovery and generative modelling. I have taken part in drug discovery competitions aimed at identifying molecules active against MurD Ligase and mapping binding sites for SARS-CoV-2 (nsp13 and nsp3) using 3D interaction fingerprints. Beyond my own research, I have been co-supervising MSc theses, worked as a Teaching Assistant for Bayesian Deep Learning Methods and Lecturer at UCL.

I am also committed to translating research into practical applications and have worked to connect academic work with industry in meaningful ways. Most recently, as an ML Research Scientist at Matterhorn Studio, I led out advanced Bayesian optimization projects in direct collaboration with Novo Nordisk. Before that, I was the ML Technical Lead at London North Eastern Railway (LNER), where I built an ISO27007-compliant AWS MLOps infrastructure and led a technical team to deploy AI models for production. My first exposure to drug discovery came was as an ML R&D Consultant for April19 Discovery, where I focused on molecular drug design for the central nervous system using generative models and multi-task DNNs.

Taking a broader view, my background is rather diverse. I have over 15 years of experience in technology, research, and finance. Earlier, I led the Data Science team at Fractal Labs, where we built a full production ML pipeline and secured a £300k InnovateUK grant. Before moving into machine learning, I built a solid base in applied mathematical modeling through visiting research positions at Imperial College London, Columbia University, and the Oxford-Man Institute. I started my career by role in quantitative finance, pricing exotic financial derivatives at KBC Global Services and consulting for PwC.

When I step away from research and mathematical modelling, my interests take a more unconventional turn. I am the creator of an artistic project that could best be described as a multi-species, multicultural alter ego. I cannot reveal my artistic identity publicly, since Banksy already has the anonymous artist space covered, but working how to sell the work while managing a steady flow of new ideas is a fascinating process in itself. I also plan to combine my creative side with my deep appreciation for canine intelligence; I am currently developing dog-friendly painting classes.