Transaction monitoring is a critical, and resource intensive, component of an effective ML programme. All too often automated monitoring systems generate an enormous volume of low value alerts - so-called false positives - which staff must sift through looking for the few that may be suspicious. This presents a great deal of risk. We can:

  • Help establish guidelines and procedures for transaction monitoring and alert investigation.

  • Review the trends in activity for banking transactions including manipulating and categorising data for investigation. 

  • Confirm whether an MT103 uses the cover method and pair the MT103 with a corresponding MT202COV.

  • Identify exceptions and deviations from SWIFT standards.

  • Advise on management information reporting systems to properly measure alert quality, track the time it takes staff to resolve alerts, and monitor alert inventory.