Soccer team in a huddle

Machine Learning at SAP: Learning Field for All

Feature Article | October 3, 2017 by Michael Zipf

The machine learning teams at SAP locations in Singapore, Walldorf, and elsewhere are bursting at the seams.

For Dirk Jendroska, head of Machine Learning Strategy and Operations at SAP, “Machine learning is a growth area for SAP and our team has an increasingly vital role to play.” The majority of SAP’s machine learning experts are driving the topic out of Singapore, Palo Alto, Brisbane, Ra’anana, Potsdam, and Walldorf.

SAP’s machine learning experts are driving the topic from locations worldwide

“But if we want to achieve our vision of making all SAP solutions intelligent,” says Jendroska, “we’ll need to draw on the expertise of everyone in Development, wherever they’re located.”

Three Ways to Implement a Machine Learning Strategy

Daniel Dahlmeier, responsible for Sales and Service in the Machine Learning team, explains how there are three ways for SAP to implement its machine learning strategy: “At the top of the pyramid are SAP’s own applications — new machine learning products like SAP Cash Application and SAP Brand Impact — that we sell as independent items on the price list.”

The level below consists of business services, built by the team and either integrated into the SAP solutions to make them intelligent or made available for purchase by customers on SAP Cloud Platform. “To take an example: SAP Service Ticket Intelligence is integrated into SAP Hybris Cloud for Service. But customers that do not run SAP Hybris Cloud for Service can also buy the business service as an API and integrate into their own application.”

The third level is SAP Leonardo Machine Learning Foundation, a portfolio that SAP presented at this year’s SAPPHIRE NOW. As SAP’s machine learning platform deployed on SAP Cloud Platform, SAP Leonardo Machine Learning Foundation offers SAP teams and customers “the technical basis on which to add machine learning logic to their solutions,” according to Dahlmeier.

Key to Success of SAP S/4HANA in the Cloud

Matthias Haendly of SAP S/4HANA Cloud Product Management describes machine learning as “vitally important,” not least for the success of SAP S/4HANA in the cloud. Because, he continues, customers expect the next generation of ERP to take process efficiency to a new level.

Machine intelligence can improve all kinds of process steps, and machine learning — like predictive analytics, to which machine learning is closely related — has the power to make these improvements happen. At the same time, it is also crucial to show customers that “SAP understands new technologies and is making them usable for its customers.” That is another reason why SAP is focusing intensely on machine learning, adds Haendly.

Phoebe Poon from the Business Development team confirms that “the topic of machine learning changes the perception of SAP and opens up new conversations with customers. Once they understand we are working in that area as well, many companies are very interested to learn about what we do with machine learning on the business side.”

“Some can’t wait to start co-innovation projects with us,” says team member Niveditha Hari. These customers are very open to new things: “They know we are reliable and that they can benefit from this more open relationship with us.”

Open to Lifelong Learning

Machine learning will impact the role of the software developer, too. According to Markus Noga, head of the Machine Learning team at the SAP Innovation Center Network, developers have always had a reputation for their open attitude to lifelong learning: “Whether it’s a new language, new operating systems, or a paradigm shift like the cloud, developers work in a fast-paced technological environment. I want to encourage every single one of them to embrace and experiment with machine learning ‒ to get a feel for what this technology can and cannot do.”

But developers are not the only ones who need machine learning expertise. It is highly relevant for the Digital Business Services area too, as business developer Christian Boos is keen to emphasize.

Colleagues in DBS need to be able to support customers that are looking to integrate machine learning business services into their solutions. By fulfilling this role, they could potentially create a feedback loop to Development “in cases where DBS or partners want to install a business service that is not yet available as a standard service on SAP Leonardo Machine Learning Foundation.”

This is how machine learning becomes a learning field for all.

Tags: , , , , , , ,

Leave a Reply