A Bold Step for Enterprise Machine Learning: These are SAP’s New Machine Learning Offerings

At SAPPHIRE NOW last year, SAP launched SAP Leonardo Machine Learning. One year later, machine learning has made another great step forward and is about to become pervasive in enterprise software.

To facilitate the fast and simple adoption of machine intelligence in the enterprise, SAP just announced a wide spectrum of new capabilities intended to provide intelligent solutions for an even broader variety of business challenges.

The intelligent enterprise framework with machine learning

The Rise of Conversational AI

Conversation is the future interface for business. That’s a trend proven by the global success of messaging apps like Facebook Messenger, WhatsApp, Kik, Telegramm, and more. Equipped with state-of-the-art chatbot support, enterprises like the French railway company SNCF use these new channels to strive for cutting-edge marketing to attract new audiences. Meanwhile, some firms struggle with managing the increasing volume of customer inquiries coming in through digital channels of all kinds and to maintain great customer service at the same time: 17 percent of people are kept on hold in lines, 42 percent speak to different agents, and 78 percent of clients end contracts because of bad support. According to Accenture, this amounts to $1.6 trillion annual losses due to poor customer service in the U.S. alone.

Following these trends and the resulting business needs, SAP just released SAP Conversational AI, now generally available, a service that includes powerful language technology and an end-to-end bot platform allowing the training, building, and monitoring of chatbots, as well as packaged bots for the telecommunication industry. Chatbots help telco businesses like SFR, France’s second largest telecommunications firm, keep customers happy by speeding up standardized processes like order tracking or SIM unlocking and relieves service agents from up to 50 percent of their repetitive workload. This allows them to focus on creative and more strategic tasks to increase overall productivity. Stay tuned: Bots for other industries will follow soon; insurance and banking are next.

RED bot created by French telecommunication firm SFR

Intelligent Applications

Our portfolio of high-value, easy-to-consume, intelligent end-to-end machine learning applications are now integrated across the SAP cloud portfolio. First, we started with back-office solutions for finance and HR. For example SAP S/4HANA Cloud comes equipped with SAP Cash Application, SAP Business Integrity Screening, and intelligent GR/IR account reconciliation. To date, the Accounts Payable service for SP Concur processed more than 2.4 million pages and invoices, and Resume Matching together with Job Standardization for SAP Fieldglass helped to handle more than 130,000 job postings and resumes. Further services like Learning Recommendations in SAP SuccessFactors solutions will be available soon. Second, also front-office solutions for marketing, sales, and service became intelligent with machine learning, such as SAP Ariba, which contains Catalogue Enrichment, as well as SAP Service Ticket Intelligence and SAP Brand Impact, which are embedded in SAP C/4HANA.

A Toolbox for Machine Learning

Machine learning is about to become a common part of enterprise software. In the past year, 79 customers implemented SAP Leonardo Machine Learning Foundation to create their own machine learning expertise in a safe, reliable, and scalable productive environment. The foundation covers the full life cycle, from raw data sets to ready-to-use applications tightly integrated into the SAP software offering. Today, it  offers 20 intelligent service APIs that you can test on the SAP API Business Hub, that function as “plug and play” and save customers business-critical resources like development or maintenance costs. The latest expansion of the offering includes brand new pre-trained services and customizable services for image and text data:

  • Customizable image object detection: Localizes and tags objects of interest in an image
  • Customizable image feature extraction: Creates a tailor-made image feature extraction service to recognize specific or fine-grained patterns in images
  • Human and face detection: Detects humans and faces in pictures and allows to find matching faces, e.g., for identification or black list check
  • Scene text recognition: Extracts text contents from any type of images, covering not only document-like images but various sceneries.

The use cases for these services are countless: A German car manufacturer used some of them to build an app that helps customers to identify a car model they see on the street with image recognition. An Austrian jewelry manufacturer automatically detects objects in images for visual product identification to optimize sales and service processes. Another scenario involves an oil and gas company that automatically extracts text from product labels in order to check if oil drums are labelled correctly.

As a next step, SAP Leonardo Machine Learning Foundation will provide comprehensive support for data scientists to collaboratively experiment and develop in a reproducible manner machine learning models for their specific use case. Along with the new ability to inspect and monitor model quality in real time during production, the foundation will become an even more complete platform for holistically managing the life cycle of self-learning software.

Car Model Detection powered by SAP Leonardo Machine Learning Foundation

AI Ethics and Society Steering Committee

Building on our activities over the past year in which we engaged with industry, policy makers, our partners, and customers to consider the impact of AI on society, SAP has now established a formal AI Ethics and Society Steering Committee. The committee is comprised of senior leaders from across the organization, including HR, Legal, Sustainability, and Machine Learning Research, who are creating and enforcing a set of guiding principles for how SAP will address the ethical and societal challenges of AI. The principles will be released in the coming months and will act as a strong basis upon which SAP will continue to expand its external and internal activities to ensure the responsible and ethical use of AI to help the world run better and improve people’s lives.

As one of the first activities, SAP has launched a global machine learning education curriculum to all its employees. The offering addresses the global workforce via a range of channels, ranging from online courses for those who are just starting to explore the area to hands-on research projects with world-leading universities for machine learning experts.

“With our virtual colleague CuRT, we automate incoming inquiries of multiple channels to deliver a higher quality and service level to our customers based on SAP Leonardo Machine Learning.”

– Pedro Ahlers, Digital Growth Manager, BASF SE

“Machine learning is the next innovation, meeting a new generation of drivers where they live, work, and play with personalized, immediate services.”

– Robert Kriehs, Senior Solution Architect, Daimler AG

If you are interested in co-innovating with SAP or are searching for direct support, contact us via email at ml@sap.com. SAPPHIRE NOW attendees can meet our SAP Leonardo Machine Learning experts onsite at booth LE514 and LE515, or join our recommended sessions.

Markus Noga is senior vice president of Machine Learning at SAP.
Katrin Schneider is communications lead for Machine Learning at SAP.