SAP Fraud Management for SAP S/4 HANA Release 1.2 SP05 was released in February 2017.
Fraud is a sub-component of Assurance and Compliance Software .
SAP Assurance and Compliance Software for SAP S/4HANA is composed of:
SAP Fraud Management for SAP S/4HANA
SAP Audit Management for SAP S/4HANA
SAP Business Partner Screening for SAP S/4HANA
SAP Tax Compliance for SAP S/4HANA
SAP Fraud Management for SAP S/4HANA analyzes business data of SAP S/4HANA on premise to detect and support the investigation of fraud and compliance problems and to help in the mitigation of fraud and compliance risks.
SAP Fraud Management for SAP S/4HANA is deployed as an add-on on SAP S/4HANA on premise and therefore shares the database schema with SAP S/4HANA. It can analyze all data in SAP S/4HANA on premise, including data from other systems that has been supplied to SAP S/4HANA by way of its business APIs, such as the Central Finance API.
Detection performance can be calibrated by what-if-analysis. The detection process and the results can be integrated into business processes and monitored (analysis of detection and investigation).
This solution can be used in any industry, including Public Sector, Banking, Health-Care, Utilities, and High-Tech.
The benefits are as follows:
Early fraud detection and integration into business processes
Quick investigation using efficient alert management
Continuous improvement of detection accuracy by minimizing false positives with real-time calibration and simulation capabilities on ultra-high data volumes
Fraud prevention using rules and predictive analytics in-memory to react to permanently changing fraud patterns
The solution enables you to create detection strategies that sift through ultra-high volumes of data for clues of fraud with rules and predictive algorithms.
Detected irregularities result in alerts indicating the anomalies found, and offer tools for further investigation, such as the Network Analysis.
Real-time business integration lets you halt and release suspicious business transactions and check new business documents while being created.
Calibration enables you to improve the automated detection by means of interactive what-if simulations.
Release 1.2 SP05 introduces the following new and changed features described below:
Intelligent Screening (New)
KPI Tiles (New)
Manage Address Screening Lists (Changed)
Manage Predictive Detection Methods (Changed)
Manage Alerts (Changed)
Detect fraud earlier and reduce loss of revenue
Decrease loss and enhance revenue protection through real-time detection of your native S4/HANA data, which helps stop fraud as early as possible.
Improve detection accuracy at a lower cost
Increase productivity of your investigation team with powerful calibration and simulation features to refine searches and improve fraud detection accuracy, even across very large volumes of data.
You can reduce the number of false-positives using what-if analyses on historical data, and using whitelisting in address screening.
Prevent and deter fraud
Use multi-rule strategies to calculate risk scores and optimize analyses of fraud scenarios.
Using information on your historical patterns, you can adapt to changing fraud patterns. You can also gain insight into which approaches are most effective in deterring fraud and make better decisions that reduce risk.
Integrate with other SAP solutions
Integrate with your organization’s other transactional and analytical software.
For example, with the integration to Payment Proposals in SAP S/4HANA, SAP Fraud Management for SAP S/4HANA can check payment proposals in SAP S/4HANA and can automatically apply blocks to suspicious payments, pending investigation.
You can use SAP Predictive Analysis to better understand fraud patterns and strategies for fraud prevention.
You can also use SAP Process Control to enhance fraud remediation using issue management workflows from the process control software. This helps you enhance ad hoc controls and policies to develop long-term improvements in fraud prevention and compliance with anti fraud regulations.
OUR CUSTOM SCENARIO;
After we installed the new version (1.2 SP05) of SAP Fraud Management Software, we created a car crash fraud demo for an insurance company with tables and views which we designed and created.
Our Claim Table is ;
Our Car information table is;
In the demo scenario we have determined our Fraud rules as follows;
If claimant age between 18 and 25 20
If car age > 15 30
If claiment have previous claim before current claim 10
İf claiment don’t have previous claim before current claim -10
If previous claim appraiser is same with current appraiser 20
We expect a claim file to be given a fraud warning if the total score is over 60.
We first created a view to use in the detection method and detection strategy definitions.
To create a view we used Eclipse IDE and SQLScript language.
SQL script and output fields of view are as shown below.
Now using transaction FRA_IMG we are starting to create and define source domain fields.
The source domain represents a business domain, which delivers detection objects to your
application system. A domain serves a set of source systems whose investigation objects and
detection objects share the same business data fields.
Define and add fields to domain .
All of the fields defined under domain.
Define Investigation and detection object types.
Our Investigation object type is;
Define Enrichment fields and keys.
Define Detection object type.
Define detection object type fields and keys.
Creating Detection Object in Fiori screen.
Start Fiori lunchpad and select Detection Method Editor tile.
Add condition fields and define condition for detection object.
Define Detection Strategy.
Add detection methods to detection strategy.
After activating Detection Strategy we can now run mass detection program.
The first accident in our example was not considered as a fraud because it did not reach enough
In the second Claim, enough score was obtained and a fraud alert was created.
Now , let’s manage our Fraud alerts.
Fraud alert detail;
You can easily see the relation between claim, claiment, appraiser, insurance agent, repair shop etc. in network analysis.
Finally we can see fraud Distribution Map.
( in our demo we have only one fraud item that's why it shows only Istanbul . )