New DBSCAN Model in Pirani: Automated and Robust Segmentation

Segmenting clients, counterparties, and other risk factors is crucial for optimizing Anti-Money Laundering (AML) prevention and ensuring regulatory compliance. At Pirani, we have enhanced our segmentation module within AML+ with a new DBSCAN model, enabling more robust, automated, and technically aligned analysis.

In this blog, we explain what DBSCAN is, how segmentation works in AML+, and how it can benefit your organization.

What is the DBSCAN Model?

DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is a clustering algorithm that identifies data groups based on density. Unlike other methods, DBSCAN allows for:

• Automatic segment identification: determines the number of clusters without needing to define them in advance.
  
  
• Number of records per segment: each group shows how many records it contains.
  
  
• Percentage distribution: understand the weight of each segment within the total dataset.
  
  
• Noise detection: atypical records are counted and reported separately.
  
  

This approach makes segmentation more precise, flexible, and reliable.

Benefits and Value for the Organization

Implementing DBSCAN in Pirani AML+ provides key technical and strategic advantages:

• Automation and efficiency: reduces manual intervention and speeds up data analysis.
  
  
• Higher accuracy and robustness: clearly identifies valid segments and noise records, ensuring reliable results.
  
  
• Complete data visibility: allows you to see the proportion of records in each segment and detect hidden patterns.
  
  
• Adaptability to different data types: works with variables such as clients, counterparties, channels, products, and jurisdictions.
  
  
• Better-informed decisions: anticipates risks and prevents suspicious activities.
  
  
• Optimized risk management: identifies high-risk groups and outliers that could affect overall analysis.
  
  
• Increased operational efficiency: automates critical processes and frees time for strategic analysis.
  
  
• Enhanced traceability and control: results are auditable and reportable, meeting regulatory requirements.
  
  
• Improved AML prevention: monitors client and counterparty behavior in a structured and reliable way.

How Segmentation Works in AML+

Segmentation in AML+ allows you to group, understand, monitor, and track the behavior and transactions of your risk factors, including clients, counterparties, channels, jurisdictions, and products.

With this functionality, your organization can:

• Prevent money laundering, terrorism financing, and proliferation of weapons of mass destruction.
  
  
• Analyze client and counterparty behavior in a structured manner.
  
  
• Manage data securely thanks to dedicated and protected infrastructure.
  
  

The module allows you to choose from different modeling techniques, including K-Means, Two-Step, and the new DBSCAN, and configure hyperparameters to achieve the most precise results based on your organization’s needs.

Conclusion

Are you already using the segmentation section?

 Available in the AML + system

Try it now! 

Don’t have the Enterprise plan?  Schedule a demo with our sales team!