Customer science is changing the way we look and anticipate the customer needs. And the Analytics space has grown quite a lot in the last few years. The analytic tools have progressed from a “sense & respond” to a “predict & act” approach! Art of prediction is no longer a mysterious science but it is well orchestrated with many tools and evolving as customer science. There are many tools and technologies in the market today that portray different techniques in doing predictive analysis!

/wp-content/uploads/2015/01/pic1_b1_626369.png

SAP as a company has come a long way and with newer technologies like predictive analytics, is leading the march. The recent acquisition of KXEN has positioned SAP in a unique way which has the combined strength of the industry and line of business solutions with home grown in-memory technology as well as the robust predictive engine. I could say this could be termed as the “perfect storm” which questions about all the aspects of handling volume, velocity, variety and veracity of data. Customers have huge sets of enterprise data from core SAP or legacy systems that need to be handled judiciously since time and data is money for the customers! With the advent of SAP HANA, data movement between ERP and HANA is addressed and customized data models have been built that handle a multitude of data sets for different purposes. Data models are also built to feed this data appropriately to the predictive engine to evangelize on the various simulations of doing predictive analysis. Yes, this is an interesting scenario that needs to be well articulated and hence the introduction of packaged solutions. The concept of packaged solutions is not new but it is engineered to provide you a proven approach while handling any real customer scenario.

/wp-content/uploads/2015/01/pic2_b1_626370.png

And yes, we do have a packaged solution for Predictive Analytics that delivers the complete end-to-end customer story and quickly builds templates that are ready to use! These best practices or guided solutions help providing quick solutions to intriguing business problems at lightning speeds! A packaged solution for about 30 different customer scenarios with out-of-the-box simulations using real enterprise data sets including state-of-the art technology like HANA and Predictive Analytics in the cloud is definitely a game changer for SAP. In the reminder of the blog, I will be referring to this packaged solution with the qualified name – Predictive Analytics Content Adoption RDS or Predictive Analytics RDS.

/wp-content/uploads/2015/01/pic3_b1_626371.png

Our updated version of the Predictive Analytics RDS is scheduled to release in Feb 2015 which commands about 30 different customer scenarios around 8 different domains (5 Industries and 3 Line of businesses). As of the current writing, the next version of the RDS is in planning and is scheduled to release in Jul 2015. With each update, we provide approaches on how to handle additional customer scenarios and different approaches of doing predictive simulations, analysis and visualizations.

You will notice how to utilize different predictive algorithms (in the context of HANA as well as SAP/non-SAP data sources) such as:

  • K-means Clustering
  • Regression or classification
  • Time series analysis
  • Association a-priori
  • Decision Tree analysis
  • Outlier detection
  • Nueral network analysis
  • Social network analysis

The different techniques also deal with a wide variety of approaches such as :

  • Predictive Analysis with HANA Predictive Analysis Library
  • Predictive Analysis with open source R
  • InfiniteInsight with InfiniteInsight Modeler
  • SQL scripting accessing HANA Predictive Analysis Library

Then comes the aspect of the different visualizations such as:

  • Predictive Analysis or Lumira UI
  • InfiniteInsight UI
  • Business Objects Explorer UI
  • UI5 (SAP HTML5)

In the next blog I will dive deeper into some of the scenarios to explain the business challenge and how it was addressed using these above predictive tools and technologies!

To report this post you need to login first.

1 Comment

You must be Logged on to comment or reply to a post.

Leave a Reply