Generating insight: training Machine Learning models with events

A German milling cutter manufacturer needed a way to automatically calculate prices for custom tools in an online shop, but predicting machine grinding times was too complex for traditional methods. By combining historical data from their event-sourced work logging system with product data from their ERP, developers trained an XGBoost machine learning model that predicts grinding times at 91% accuracy, enabling fully automated, on-demand price calculations without specialized staff.