Rule-based risk scoring (0–5 points) using the same logic as the Hive data transformations in our analytics pipeline