Segmentation allows you to group, properly understand and track the behavior and information of your customers, counterparties, channels, jurisdictions and products, helping to strengthen monitoring and establish patterns for risk analysis.
With Pirani's new segmentation section you will be able to prevent money laundering in your organization by grouping and understanding the characteristics of your risk factors in order to monitor and follow up on their behavior and transactions to ensure that they are not involved in suspicious operations that promote money laundering, terrorist financing and proliferation of weapons of mass destruction. This new section has a dedicated infrastructure that allows you to store the information independently and isolated to give it the care that the nature of the information needs.
You will be able to separate your data into sets by means of the "data set partition", this functionality will allow you to distinguish the data by means of groups to make segmentation easier and more optimal. For example, you can divide your customers into two sets, natural persons and legal entities in order to make a proper segmentation and monitoring since they do not have the same behavior.
In the process of creating your segmentation model you will be able to choose the variables with which you want to study your datasets, as well as check that the variables have the appropriate population for the segmentation to be of quality.
Discover how simple it is to manage your LAFT risks and create segmentation models with Pirani. Try our tool FREE and tell me in the comments of the article what you think about our segmentation section.
For the segmentation process the tool uses a data mining process consisting of 4 steps that guarantee the quality of the data: first the data is converted into numerical data, then the outliers are calculated using the criterion of 3 times the population standard deviation, then the data is scaled and finally the dimensionality of the dataset is reduced for a better response in the processing.
For the data modeling process you can choose the modeling technique to be used between K-Means and Two-stage, and also choose the method of calculating the optimal number of segments between Elbow and Silhouette.
From the variables determined in the segmentation process you will obtain a report with the characterization of each of the segments, this is given by means of a clear description of the characteristics that each group has, which will help me to monitor and follow up the behavior of the segments.
In this article I told you the main features of our segmentation section but if you want to know in more depth how to create your segmentation models I leave you this tutorial where I explain in detail everything about this topic ;).
⚡ Remember that to have this new section you must have the AML+ management system.