Optimize your financial crime analytics and automation

Data analytics involves analyzing raw data and then drawing conclusions and generating insights based on the results of the analysis. This broad, simple definition provides a big umbrella that covers a diverse range of techniques and processes.

Many people think of analytics as an inherently technology-enabled process. But technically, a person could perform simple data analytics techniques with just pen and paper. And many financial crime programs are already applying data analytics manually via spreadsheets. A financial crime analyst looks at a table, analyzes data, creates pivot tables or otherwise manipulates the data, and comes away with new information – that’s financial crime analytics.

However, manual analytics processes require a lot of time and management from humans. As an organization grows, data sets get bigger, and problems become more complex. In the face of these new challenges, manual analytics processes scale up in a sustainable and consistent way. That’s where automation comes in.

Automation can take place both before and after financial crime analytics processes to add efficiency and value at each end. For example, consider a financial services company whose financial crime leadership team wants to implement their process for model optimization so they have a better idea of when and what to calibrate.

Right now, the organization only performs calibration once per year as dictated by its calibration methodology. The financial crime team wants to conduct these calibration exercises on a more responsive, risk-based basis.

How can the team accomplish this? Here’s an example of a potential process:

1. Based on the defined goal, the organization could implement data analytics software with automation capabilities.
2. A bot can automate the process of gathering data from anti-money laundering transaction monitoring systems, then automatically perform data analytics calculations that provide metrics about transaction monitoring scenario performance. These metrics can include data such as the total number of generated alerts, the percentage of alerts that resulted in a case, the percentage of cases that resulted in a suspicious activity report, and the date when the threshold for each scenario was last adjusted.
3. Based on the calculations, the software can generate automatic reporting that provides a summary and visualizations of the results.
4. By examining the changes in the metrics, the team can choose which transaction monitoring scenarios should receive above-the-line and below-the-line calibration. The software can provide a notification when a change in the calculations occur so that the team can plan its next calibration exercise.

The organization now has turned a static process into one that’s risk-based and responsive. Financial crime analytics provides the insight and ability to schedule calibration in response to risks and trends, while automation makes the process consistent, sustainable, and scalable.