Fraud detection: Point of Compromise

Business challenge

  • Credit card network needs to identify common points of compromise for fraud cases
  • Current analysis is based on simple, human-selected criteria; only a subset of the available features is used


  • An advanced machine learning system that utilizes subtle structures in the data to give faster and more accurate PoC detection

Business benefit

  • Faster and more accurate detection of compromised retailers allows faster and more accurate detection of other compromised cards
  • Once a PoC is identified, investigations are triggered to identify the perpetrators