At Betera, I replaced external CRM segmentation tooling with an in-house engine combining RFM, machine learning, and rules-based routing. It reduced licensing expenditure by about 21% year over year and lifted retention 29% across target campaigns.
I also designed Betera’s automated deposit and cashflow forecasting suite, reducing MAPE by 31% against expert forecasts. I’m currently co-architecting the migration of more than 10 models to a vendor-agnostic stack, with migration stages delivered without pipeline downtime.
Before Betera, I developed state-space models, Kalman filters, and Monte Carlo simulations for telecom and satellite systems, and translated mathematical models into production C++ for autonomous control systems. I bring quantitative modeling together with production ML systems and business-focused analytics.

