Automated Exceptions Handling Engine leverages semi-supervised learning to understand & learn from previous reconciliation cases and then predict both known & unknown exception cases without human intervention. It resolves exceptions by improving break classification according to clients’ requirements and automatically builds exceptions models from historical data. 

RS is targeted to reduce exception investigation & resolution time by 50-75% and achieve over 95% accuracy in exception detection & adjustment amount prediction.