Worked at the interface of the AI and biotech teams on structure-based drug discovery, generative molecular design, and applied machine learning — the link between the computational tools and the biology. Contributing author on a NeurIPS 2025 publication on uncertainty quantification in deep regression.

• Built a modular synthetic-data-generation workflow orchestrating GNINA, DiffDock and AlphaFold 2; sourced datasets and handled inter-tool format conversion
• Predicted protein structures and binding affinities for protein complexes, including metal-ion and ATP cofactors, using Boltz, AlphaFold, Foldseek and BioEMU
• On the engine team of Uncharted, an internal venture building a multi-objective small-molecule optimisation platform; developed the binding-affinity objective
• Analysed metagenomic data, prototyped a novel operon-design approach, and ran preclinical power analyses
• On the NeurIPS 2025 work: ran empirical validation and fine-tuned/retrained the model in PyTorch, informed by a Bayesian / Monte Carlo perspective
• Ran GPU workloads on an NVIDIA DGX cluster and AWS; Docker, Git and Azure DevOps within an Agile/Scrum team

Quantitative sensory biology within an international Human Frontier Science Program (HFSP) collaboration, with emphasis on statistical modelling and signal processing of behavioural data.

• Analysed experimental data using Bayesian multilevel regression models in R and Stan
• Designed and executed behavioural experiments with automated data acquisition pipelines
• Contributed to interdisciplinary publications spanning biology, optics, and computation

Computational and statistical analysis of visual systems in invertebrates, combining experimental biology with quantitative modelling.

• Built statistical models using Bayesian multilevel regression in R and Stan; additional analysis in Python and Matlab
• Developed optical and geometric models to estimate spatial resolution theoretically
• Designed novel experimental protocols for quantifying animal visual performance