Merchandising teams watch sales, orders, and commissions week over week. When something drifts in a bad way, you want a signal early. You also want fewer false alarms so people still trust the system. Our alerting stack pulls structured performance data, prepares slices for different kinds of …

Speech therapists in France and Switzerland often live inside Word: session notes, initial assessments, renewal reports, school coordination. The writing is high stakes (tone, structure, traceability) and high friction (hours reformatting the same clinical skeleton). Cloud assistants are a hard sell …

Municipal and cantonal open geodata is powerful and noisy. REST feature services expose dozens of layers, attribute names drift, and row counts sit in the hundreds of thousands. Analysts who need one clean building-level table for dashboards do not want to click through ArcGIS every week. They want …

Museum catalogs live or die on consistent metadata. For a large LP collection in ethnomusicology, that metadata often starts on the object itself: tiny type, multiple languages, stickers, and decades of wear. Typing that by hand does not scale. In the field, the pain is blunt: teams reported …

Small and mid-sized teams do not lack data. They lack a safe, fast path from a plain English question to a correct aggregate or filter over their tables. Spreadsheets scale poorly. A full BI rollout is heavy. In the early wave of LLM tooling, the obvious experiment was Text-to-SQL: let the model …

If you study algorithmic trading, you quickly hit a tension. Live markets are expensive to replay at full fidelity, and toy models (pure Poisson arrivals, IID noise) look nothing like the bursty, directional microstructure you see in real message feeds. In 2024, our group set out to build a market …