Work
A selection of applied AI, machine-learning and product-engineering projects, mixing client engagements with self-directed research. Each links through to a fuller write-up.
Chat:R, an enterprise language-model platform
Enterprise LLM platform · React, Node.js, MongoDB · Azure · 2025-present
An enterprise language-model platform serving roughly 1,000 staff at a European public-affairs consultancy, with custom Projects, agent-linking and SharePoint MCP gateway features, deployed across three Azure environments.
Cell-spheroid segmentation for high-throughput microscopy
Deep-learning computer vision · U-Net, PyTorch · clinical research
A plain U-Net segmentation model for high-throughput cell-spheroid microscopy, built as research engineering for a surgical-biotechnology group, reaching a validation intersection-over-union of 0.968 and Dice coefficient of 0.983.
London Property Finder, applied machine learning
Applied ML · XGBoost, quantile regression, geospatial · self-directed research
An interactive single-page tool that scores live London property listings against a stacked machine-learning ensemble trained on 2.18 million Land Registry transactions, to test whether public data alone can identify mispriced homes.