Projects
Selected work across the last few years.
-
Prompt to Audience
A modular agentic RAG pipeline that translates plain-English queries into complex audience trees for a Big Six advertising agency.
-
Agentic Company Discovery
An in-house pipeline that bootstraps a client-ready sample of companies for a given sector, using LLM labellers and SetFit classifiers on top of glass.ai's web-scraped data.
-
Custom Loss Functions for Low-Label Classification
Two production classifiers at glass.ai where the interesting work was the loss function - a genetic-algorithm-searched adversarial PU loss, and DPO adapted to small transformers for pairwise preference tasks.
-
Electoral Prediction from Social Media
A seat-by-seat model of the 2019 UK General Election from Twitter, combining probabilistic constituency geolocation, three parallel methods of inferring a user's partisan leaning, and a simulated-annealing bias correction to bridge the gap between the Twittersphere and the electorate.