The New Machine Learning AML Platform

The Challenge

As financial crimes grow in complexity and scale, financial institutions are facing increasing responsibilities and costs that come with the need to comply with Anti-Money Laundering (AML) regulations. With current AML monitoring systems unable to keep up with rapidly shifting technologies, analysts are caught in time-consuming, manual processes that can take up to a week for particularly complex transactions.

The Solution

Introducing a Machine Learning AML Transaction monitoring system.

Lower costs, higher detection efficiency
Co-developed by OCBC Group Legal and Compliance team and fintech company - Thetaray and empowered by The Open Vault at OCBC team, this system uses Artificial Intelligence and Machine Learning technologies to dramatically reduce the costs of responding to changing regulatory requirements. The most significant advantage ThetaRay has over existing systems is that it can identify new threats and risks just as they are forming – vastly increasing efficiency and detection rates. A combination of Machine Learning and Data Analytics allows ThetaRay to evolve with different scenarios by “learning” from new use cases and redefining parameters based on contextual indicators.

Innovating with passion
The project began life with just three members from the OCBC Group Legal Risk and Compliance team. Determined to see their passion project through despite its complexity, the team even committed time outside of their day-to-day work to ensure its successful implementation.

The Results

Efficiency Gained

Reduction of False Positive Alerts

Early results have shown a 4.5x efficiency gain, an increase of AML detection mechanism and a 35% reduction in the number of alerts that did not require further review. This was achieved by categorising more effectively flagged transactions by their risk levels.