Graduate capstones and applied studies: real problems taken from raw data to a working result. Click any one to read the approach in full.
Semester-long theses that take a real problem all the way through, and the most rigorous work we've done.
Topic modeling and LLM analysis across customer-agent conversations, built into an end-to-end view of how every agent performs: what customers raise, how it's handled, and where coaching pays off.
Read → Graduate capstoneAutoencoder models that let satellites not built for the job detect space weather anomalies, squeezing scientific value from hardware already flying.
Read →Focused, self-directed projects that answer a specific business question from data to decision.
Forecasting component failures and scheduling maintenance before downtime hits, instead of after.
Read → Applied studyLanguage models that surface recovery opportunities buried in claim files: work that's tedious for people and cheap for machines.
Read → Applied studyBehavioral segmentation across a retail catalog: who buys, why they buy, and what to do differently about each group.
Read →Work delivered inside partner organizations, on their problems and their data.
Rowan is at Arcfield working in defense systems engineering, Jonah is in Virginia Tech's Institutional Data office, and Logan is consulting at Forvis Mazars. Write-ups will appear here as that work wraps and can be shared, same honest treatment as everything else on this page. Want your organization on this list? Start a conversation →
Bring us what's slowing you down. The first session is free.