success story

Combating money laundering with AI-based advanced analytics for a Spanish bank

Virtusa successfully helped a leading Spanish bank fight financial fraud with analytics-based anti-money laundering (AML) solution.

The client is a Spain-based multinational financial services company that maintains a presence in all global financial centers as one of the largest banking institutions in the world. The bank’s analytics and compliance team are responsible for tracking money laundering schemes across its corporate clients.

The Spanish financial regulations force all banks to have an established rule-based monitoring system to track several money-laundering transactions and report suspicious transactions to the regulators. However, with the surge in the sophistication of money laundering schemes, the rule-based system is becoming obsolete and ineffective. Hence, the need to use more advanced analytics is inevitable.

The Challenge

Being a regulated bank, the client uses an established rule-based monitoring system to track money-laundering transactions and report suspicious transactions to the regulators.

However, with the rising sophistication in money laundering schemes, the rule-based system is becoming obsolete and ineffective. As a result, the client struggled to infer some of the money laundering and suspicious transaction incidents. The bank embarked on a compliance program to building AI patterns for transaction monitoring and payments screening across various business domains and products within its corporate banking business. So, over six months, it deployed 5-6 patterns across three products. However, while developing and deploying these patterns, the bank faced some additional challenges: 

  • Many false positives resulting from the limitation of a one-sided rule-based AML engine
  • Unavailable alert evaluation scoring system resulting in shallow visibility on the significance of an alert
  • Rising need to build an advanced analytics engine while supporting the rule-based engine simultaneously
  • Increased perceived data quality challenges that had a direct bearing on the quality of alerts generated
The Solution

Mapping the background and identified challenges to the client’s business context, Virtusa deployed its data science and data engineering team to build AI models to help it leverage the benefits of AI/AA

Anti-Money Laundering (AML) Solutions - Key features
  • Added an analytics-based engine with multiple money laundering patterns to improve the quality of suspicious alert triggers using outlier detection techniques
  • Built a clustering model from scratch to appropriately classify the client’s corporate clients and conduct a peers-behavior analysis to fortify the analytics engine
  • Implemented an alert scoring mechanism to improve the triage system by using the term frequency-inverse document frequency technique
  • Pattern refactorization and optimization to improve performance and reduce production costs

 

The Benefits

Together with the collaborative efforts of the client’s project team, we achieved the following results:

  • 25% reduction in false positives of alert triggers with a robust engine 
  • Improved productivity saved 20 hours per week with consistent identification of relevant alerts
  • Optimized cost savings
  • Reduced investigation time by 50%
Analytics, Insights, Data

Modernize your data platforms and apply AI/ML to redefine, re-engineer business processes leading to superior customer experience and higher productivity

Related content