Technology Blogs by SAP
Learn how to extend and personalize SAP applications. Follow the SAP technology blog for insights into SAP BTP, ABAP, SAP Analytics Cloud, SAP HANA, and more.
Showing results for 
Search instead for 
Did you mean: 
Product and Topic Expert
Product and Topic Expert
Thanks to Dr. maria.fay2#overview (main contributor) for drafting this post together with me.

What machine learning and artificial intelligence mean for me as an enterprise?  When and how to start? What is the impact of machine learning on each business domain? These are the questions that SAP addressed at the SAP AI Business Services Focus Week event at Mode-2 Garage Switzerland. 

Why: Raising awareness and building bridges 

The event focused on two key areas: the enablement on the embedded machine learning capabilities and applications in S/4HANA, C/4HANA and SAP SuccessFactors, as well as hands-on sessions around generic SAP AI Business Services.  

While we see customers show interest in AI capabilities of SAP, among the areas still interesting them most of all are the core business processes and such domains as finance, procurement, or sales. In the meantime, there have been over 30 machine learning and predictive analytics scenarios delivered for S/4HANA – that can be deployed via Predictive Analytics Integrator (PAi) connector or over the SAP Cloud Platform (see the Architectural Guidance document) – and dozens more are in the pipeline. There are already several channels to bring those scenarios closer to the customers – development organization is leading co-innovation projectsExplore services for intelligent ERP is a consulting offering to explore innovation scenarios for and beyond S/4HANA. So, the Focus Week event has allowed to strengthen these activities by offering a full day of sessions around intelligent scenarios for each line-of-business, with interactive demos in the system, to demystify those scenarios and get instant customer feedback on the relevance of those cases for their business. 

The second focus block of the event were the SAP AI Business Services. These re-usable microservices are available via the SAP Cloud Platform. They are tangible machine learning capabilities that provide a generic business functionality to automate and optimize business processes. Thanks to the modular and API-based architecture they can easily be integrated across various processes and solutions. In that way, they enrich the customer experience and provide immediate business value. E.g. if you want to unlock the unstructured data from a PDF invoice and see the results in your ERP, the Data Information Extraction service will perform the extraction, and return the data via a REST API to the target system. The value-based pricing model makes the usage of SAP AI Business Services easy to understand and to consume.  

Apart of highlighting the value of each application and service the event aimed at identifying and qualifying potential use cases in customer business setting. The event incorporated the respective overview sessions followed up by three days of rapid prototyping. With support of the SAP team (development + consulting) three customers designed the use cases and potential target architectures. Not only they performed hands-on exercises by connecting to the services via API and Postman, but they also uploaded previously prepared data sets and tested the services in real-time.  

What: SAP AI Business Services 

Let's have a short deep-dive and see what business value these services actually provide. 

Data Attribute Recommendation applies machine learning to match and classify data records automatically, making more efficient and higher-quality master data management a reality. 

When to use? Predicting: material classes & material characteristics; commodity codes; product records/hierarchies. 

Fig 1: Example of a use case for Data Attribute Recommendation. Source: SAP Internal - AI Business Services (2019).


Document Information Extraction is applying several machine learning models to digitalize unstructured business documents by extracting structured semantic information and matching this information with relevant business data. The service requires a PDF document as input, and returns the information from header fields and tables in JSON format. 

When to use? extraction of data from: invoices; payment advices; sales orders; goods receipts; … 

Fig 2: Example of a use case for Document Information Extraction. Source: SAP Internal - AI Business Services (2019).


Service Ticket Intelligence automatically determines ticket categories and proposes problem resolutions based on similar tickets.  

When to use? IT, HR and customer support – processing: social media posts, e-mails, website queries. 

Fig 3: Example of a use case for Service Ticket Intelligence. Source: SAP Internal - AI Business Services (2019).


How: Focus Weeks

Now that we know when and what to use, let us see how the Focus Week format addressed those services.

Focus Week format is a logical evolution of existing Mode-2 Garage Switzerland formats. Over the past year Mode-2 Garage has been successfully hosting ‘One Week – One Customer - One Challenge’ rapid prototyping events for individual customers. As opposed to the regular approach, where the customer provides the design and prototype challenge, the newly introduced Focus Week determines the challenge frame in form of a specific focus topic. In that way multiple customers can come together and receive the enablement and support directly from SAP product experts, development teams, and fellow consultants. They jointly learn and experience the latest solutions and capabilities in a hands-on manner. On top they scout and define their business-relevant use case, they can apply the latest technology to. As the feedback showed, the variety of use cases discussed at the event, the level of content, and interactive nature of the sessions with demos and services in action were largely appreciated by the customers.

What’s next?

Stay tuned for the upcoming hackathons on SAP AI Business Services or contact us directly:

  • Tomasz:

  • Maria:

Why is it worth it? Well, as they put it, anything that is a process can and will be run by AI ?.