Thursday, April 16, 2026
Jose Borges
Forecasting Store Sales at Scale: Lessons from 1,782 Time Series
How I cut forecast error by 38% on a Kaggle retail dataset - and what it taught me about feature engineering, validation design, and when boosted trees beat deep learning.
Tuesday, February 10, 2026
Jose Borges
Predicting Churn with Machine Learning
How we built a churn prediction model that identifies at-risk customers before they leave.