SAP’s Predictive capabilities – applicability to SAP High Performance Applications.

Loads of emails/every next blog/numerous tech gurus, they all talk of transforming the static behaviors and catching the trends dynamically – yes they are talking about Predictive.

Predictive capabilities have been a whisper of the lip over the past couple of years, and where does SAP stands – I will try to unleash the power and applicability of predictive in SAP’s applications.

Predictive Analysis as the name goes, is a model that allows you to intuitively design complex predictive models based on current and historical data within the enterprise, to depict trends and share hidden insights of activities performed on that data. SAP Predictive Analysis delivers the value by leveraging the capabilities and power of Big Data with SAP HANA. It’s an intuitive, drag-and-drop, code-free experience with enough power for data scientists to conduct more sophisticated analysis using Big Data, yet simple enough to allow business analysts to conduct forward-looking analysis using departmental data from Excel.

Various modes of predictive capabilities exist, to get the data and provide visualization to this prediction. Some key highlights which brings the capabilities for user:

• Intuitively design complex predictive models
• Visualize, discover, and share hidden insights
• Unleash Big Data with SAP HANA’s power
• Embed Predictive in to Apps and BI environments
• Real-time answers

Figure 1.1 details on the high level architecture of where PA tools fit to a typical SAP solution.

Predictive Libraries are available in the following flavors:

• AFL (Application Function Library): This resides inside of HANA DB. As you see in Figure 1.1, these libraries are available as a part of HANA database engine. Check for your HANA server version and install the AFL plug-in. AFL includes BFL (Business Function Library) & PAL (Predictive Analysis Libraries). (For AFL, click here).
• BOBJ Predictive Analytics: BOBJ predictive analysis can be integrated to leverage the data sitting in HANA to derive patterns and update the content for HPAs. BOBJ analytic tools have predictive capabilities to build models and generate visualization for your dataset.

• SAP Predictive Analysis solution can also be used to integrate and leverage the HANA predictive libraries, once HANA is connected in online mode.This integration serves the user friendly SQL procedures which can be easily consumed by the end product, like any HANA based HPA in this case.(Click here for more details on SAP Predictive to HANA PAL integration.)

Consuming the predictive capabilities to generate the functional logic in the SAP’s High Performance Applications.

SAP’s HPAs come with native HANA architecture, they are installed and scale on HANA as database. As with Figure 1.2, the data from various sources is collected into the HANA DB via SAP LT replicators or via BOBJ dataservices for SAP or non-SAP systems respectively (figure 1.1 for details).Further, these libraries within AFL can be used to build wrapper procedures which are the actual predictive models based on any historical records brought from various databases.
These predictive models can further be called within the execution procedures within the application.

(How to deploy SAP HANA predictive analysis libraries? Click here for step-by-step approach.)

SAP PA tool can be defined to use these predictive models to understand the behavior of users within an enterprise data. Based on historical trends and the actions exhibited by a typical business user, models and procedures can be predicted by these algorithms.

Example Use Case: SAP Fraud Management – Integration with Predictive Analytics

With SAP’s portfolio of

HPAs, we have applications like SAP Fraud Management are live applications using such capabilities to detect frauds by continuously refining and updating the rules, based on the predictive models detected by the analysis based on the current and historical data. It provides the following:

• Integrates with Predictive Analytic tools to deliver easy to consume models in form of detection strategies.
• Combines complex mathematical and statistical analysis, to optimize based on ever changing data condition.
• Seamless integration between HANA’s Predictive Analytics Library and SAP Predictive Analysis, enabling access to a rich library of algorithm and visualization tools.

Advantage

• Define better strategies which ensures the most relevant and critical issues are addressed.
• Provides insight to reduce false positives and drive improvements with ever changing patterns for fraud to happen.

More details on Fraud Management are available on help.sap.com/fra.

Similarly with above defined approach, we can use predictive capabilities to innovate with SAP/non-SAP solutions in today’s industry dynamics.

Important links are tagged.