Client Introduction
The client is a leading US-based super-regional bank with over USD 200 billion in asset size, offering retail and commercial banking services, wealth management, investment solutions, and more.
Problem Statement
The bank was facing an increasing number of regulatory sanctions leading to increased false positives due to:
- Heightened geopolitical risks.
- Increased regulatory scrutiny.
Business Impact
- Onboarding delays for commercial and retail customers.
- Data inconsistencies in the system.
- Increased manual efforts to conduct ongoing KYC.
- Escalated compliance costs.
- Degraded customer experience.
Key Requirements
- Data transmission from WLM to solution and decision alerts from solution back to WLM.
- Reduction in manual or human intervention.
- Real-time processing of alerts.
- Generation of audit trails and reports.
Solution Approach
Anaptyss Financial Crime Compliance experts assessed the case, determining key intervention areas and the remedial approach as follows:
- False alerts for negative news and Politically Exposed Persons (PEPs).
- Domestic customers based in the United States.
- Individual/retail customers.
They determined that an advanced “logic-driven and evolutionary” algorithmic automation solution could reduce the false positives significantly. A solution based on this technique could drastically reduce the manual efforts, risks, and regulatory scrutiny.
ALFA: AI/ML-Powered Enterprise-Grade AML Compliance Solution
Anaptyss deployed ALFA (Automated Learning for Financial Alerts) — a proprietary digital solution that enables real-time transaction monitoring, watchlist screening, KYC risk profiling, and alerts investigation with high accuracy.
ALFA addressed the problem as follows:
- Deployed rule-based semantic logic for date of birth, year of birth, address, and location.
- Built a decision tree to generate an audit trail comprising documentary attachments with date-time stamps.
- Categorized the risks based on the type of watchlist.
- Offered “out-of-the-box” scalability to allow the mapping of additional rules based on user/machine-defined inputs.
Business Outcomes
- 75% accuracy for detecting false positives.
- Significant reduction in manual efforts as only 25% of genuinely suspicious and complicated cases were remaining for Level 1 and Level 2 reviews.
- Improved compliance with regulatory sanctions.
- Effective safeguard against financial crimes and geopolitical risks.
Want to learn more about ALFA?
Write to us: info@anaptyss.com