I am a machine learning engineer and data scientist who builds data platforms and the AI systems that run on top of them. I hold a PhD in Computer Science from Georgetown, where my research modernized topic models to handle the noise, time, and sparsity of real social media text. That work became the topic-noise model family, published at the ACM Web Conference, PAKDD, and Knowledge and Information Systems, alongside a widely cited survey, "The Evolution of Topic Modeling," in ACM Computing Surveys.

Since 2022 I have been Head of Data Science at 5-Out, directing forecasting and ML for restaurant operators: the models, the monitoring, and the evaluation methodology that decides whether a forecast is trustworthy. I also run Churchill Software, a small senior consultancy that takes clients from prototype to production, with recent work spanning a commercial construction takeoff product and a pipeline that turns messy financial statements into a reconciled canonical schema. I work across OpenAI and Anthropic APIs, open-weight models, SQL and NoSQL, ML pipelines, and AWS, and I care most about the gap between a demo that works once and a system that keeps working.