In the strict regulatory environment of AML detection Banks are forced to employ broad sweeping detection models that have high detection rates but also extremely high false positive rates, creating a huge financial burden for investigation.

spotixx proposes a unique software solution that utilizes machine learning algorithms to rank AML alerts to ultimately lead to a reduction in false positive rates due to alert hibernation. We create human readable business rules that directly target filed cases for selected detection and false positive rates. These rules are discovered through a process of supervised machine learning and can adapt and evolve to changing criminal behaviors in the Bank’s customer population.
As the spotixx business rules are platform agnostic and easily interpretable by technical and non-technical people alike they can be used for a broad set of applications.

Use cases are:

  • Input to an alert scoring /post-processing model
  • Input to an alert hibernation model
  • Identification or optimization of risk factors
  • Threshold tuning
  • Identification of new detection rules

 

More information:

pdficon