Tookitaki AMLS Transaction Monitoring

by Tookitaki Holding Pte. Ltd

An award-winning machine learning solution that detects false negatives and reduces false positives.

Tookitaki Transaction Monitoring solution is a robust money laundering detection and alerts management module within our Anti-Money Laundering Suite (AMLS). Built on distributed systems, it leverages advanced machine learning techniques like AutoML, federated learning and network science to accurately detect complex money laundering activities and triage legacy system alerts into high, medium and low-risk buckets with a clear, auditable explanation ensuring higher AML risk coverage and faster alerts disposition. AMLS TM uses the main techniques of supervised and unsupervised models to generate predictions for secondary scoring of alerts and anomalies. The models use the already generated alerts from the current system like Actimize-SAM and feeds it to the supervised model to prioritize the alerts into high, medium, low risk and the raw data followed by building new features on it and use unsupervised models like dynamic clustering as models for determining anomalies finally resulting in reducing the false positives and increasing the true positives.

Quantifying business value of AMLS TM Module:
  • 40-50% reduction in false alerts with 85-90% SARs in high risk bucket
  • 3-5% uplift in overall false negatives or missed true cases
  • 70% faster time to market as against other AI solutions

At a glance