I’m a data scientist with 9 years of experience building probabilistic models, and the validation that makes them worth trusting. I specialise in Bayesian inference, uncertainty quantification and model validation, mostly for problems where a point estimate would be actively misleading: expensive simulators, data that’s sparse or noisy or both, hypotheses that need something more rigorous than eyeballing to tell apart.
Recent work includes building a Bayesian inference pipeline that helped achieve world-first 5 nanometre resolution in single-molecule localisation microscopy, a Monte Carlo model pricing a commercial siting decision under uncertainty, and exploring whether purely machine learning approaches can reconstruct historical proto-languages from their modern descendants.