SAP AG has released a new predictive analytics engine – the first solution to come out of its acquisition of KXEN Inc., a U.S.-based predictive analytics software provider, which took place earlier this fall.

Branded as “InfiniteInsight,” the same name as one of KXEN’s flagship products, SAP’s new solution provides a raft of features, including analytics data sets, which allow for predictive modeling. Using something the “explorer feature,” users can define a set of analytical records, which are constantly applied over and over again to create a predictive model.

(Image: SAP). Screenshot of InfiniteInsight, with a model output showing contributions by variable.
(Image: SAP). Screenshot of InfiniteInsight, with a model output showing contributions by variable.

Automating the process means that data scientists don’t have to model everything individually.
The InfiniteInsight solution also provides accurate models, a feature that automatically builds models for data mining. If a user has a specific business problem he or she wants to solve, this feature can generate an algorithm almost instantly.

SAP and KXEN are also providing actionable scores, which can be integrated with call centres and websites. For SAP customers who are already using SAP Hana, or for customers who have their own database platforms, the scoring feature can tap into their databases to provide optimized scoring equations.

And with the peak performance feature, InfiniteInsight allows users schedule model refreshes as often as they need to, something that comes up when users add a large number of campaigns, or when they make changes to the model when they need to target small groups of people.

By launching InfiniteInsights along with KXEN, SAP is trying to get more business users on board with analyzing their data – something that’s typically more in the realm of data scientists, said James Fisher, SAP’s vice-president of marketing for analytics solutions. He spoke to ITBusiness.ca in an interview last week.

“There’s a real sense that organizations are really struggling with some of the traditional ways predictive analytics have been deployed – a kind of very highly manual type of process, very time-consuming, handcrafted, repetitive methods to build models, which require this very rich data design skills,” he said.

“And while we’re not advocating for one second that these data scientists are not important … we think the technology needs to evolve in terms of the automation of a lot of this creation of predictive models.”

InfiniteInsight is now available worldwide, with pricing available upon request.

Share on LinkedIn Comment on this article Share with Google+
Around the Web
More Articles