At Aleatica, I designed a traffic-demand forecasting pipeline for predicting demand 72 hours ahead. My final stacking ensemble achieved an MAE of 90.0 and an R of 0.85; validated on historical data, it would raise staff allocation accuracy from 41% to 65%.
At Alumigraf, I built Power BI dashboards used daily by the CEO and teams across the business, and replaced paper-based non-conformance tracking with a digital system in SAP Business One. Earlier, at JG Ingenieros, I analysed maintenance work orders in R and authored a technical white paper with findings and recommendations.
