SAP is conducting the “Business meets AI” hackathon format at several locations. Companies develop AI use cases into a functioning minimum viable product (MVP) in five days.

This is how AI hackathons bring ideas to the MVP

What slows down many companies when introducing artificial intelligence is rarely for technical reasons. It's a lack of time, unclear responsibilities, a lack of focus and the difficulty of having the right people from business and IT at the same table at the same time. This is exactly where SAP’s “Business meets AI” hackathon format comes into play. This year it took place in parallel at several locations for the first time: in Zurich, Munich, Berlin and, at a later date, in Vienna. The approach is the same at all locations.

Five days, an individual use case, a working MVP

The format is aimed at companies that have a specific use case and would like to implement it in five days together with SAP experts. The end result is a functioning MVP.

Participation requires an application. This year there were 30 applications for nine places in Zurich, and almost ten teams with over 40 participants took part in Munich. Use cases are evaluated by a jury based on whether they can be implemented within the time frame, whether they have a measurable business value and whether they are not already covered by standard functionality. Alexander Finger, CTO at SAP Switzerland and one of the initiators of the format, describes the claim as follows: “The use case should be a little too much for a Friday afternoon and a little too little for a real project.”

The preparation begins before the actual event. System access, data availability and authorizations are clarified in advance so that teams can work productively from day one. The customer teams should be interdisciplinary and consist of roles such as business users, process owners, admins and developers. You will be accompanied by a dedicated SAP coach who will bring in additional specialists if necessary. These coaches were cited across locations as a key success factor.

Results from Zurich and Munich (and Berlin?)

Functional MVPs that address real company problems were created at the locations this year.

In Zurich Four teams received awards in the Real World Impact, Technology Factor, Wow Factor and CTO Award categories – with solutions in the areas of tax compliance, maintenance, procurement and contract processing.

In Munich (Garching), the teams addressed use cases from HR document processing, sales automation and accounts payable. Here, too, four use cases were honored by an expert jury. A description from the perspective of the teams of selected winners will appear in separate SAPNews features.

In Berlin AI application areas were not only discussed theoretically with organizations from the public sector, but also tested very pragmatically. Based on the use cases presented, executable applications were developed within a very short time. These solve some procedural problems that are years old and digitize the corresponding processes. This involved, among other things, the automatic configuration of SAP systems based on concepts, the efficient evaluation and processing of invoices and an effective way to process dispute cases.

Kay Kretschmer, Coach Lead in Garching, explained: “The combination of customer ideas, technical and business AI expertise of the SAP coaches and the teamwork between the customers and SAP clearly shows that the added value of Business AI can be realized particularly when the focus is on the business value and not the technology. AI for AI's sake is usually doomed to failure.”

What can be seen across locations: Teams report that they achieve more in one hackathon week than in six months of regular project work. The reason lies in the structure and the focus on a clearly defined scope with business and IT in the same room with quick, short coordination.

The results show the diversity in the application of AI, even if there are patterns. Some applications are small and special, but you can always see the added value that AI offers. In Munich, one team was able to reduce the processing time per tender process from three to one hour, and another reduced the input steps in master data maintenance by 50 percent.

Technological building blocks

Recurring technology components have emerged across all locations and use cases. SAP AI Core and the Generative AI Hub form the backbone for interacting with AI models. SAP Document AI is used for structured extraction from unstructured documents. SAP Build Process Automation takes care of background processing and workflow control. The Cloud Application Programming Model (CAP) is used for backend services, often in combination with existing SAP system connections.

A format that scales

The format was developed and iterated in Switzerland over several years before being rolled out to the region. Claas Hoffmann, hackathon host at SAP Switzerland and part of the organizational team from the start, describes the development: “Every time it is a new level of customer cases, as well as in terms of preparation and implementation.”

There is now a cross-location alumni community in which participants exchange experiences and report on the progress of their cases. The hackathon is therefore not designed as a complete event, but rather as a stop on an implementation journey. Therefore, all participants are winners. Everyone has gained experience and knowledge. Of course, it is now important to transfer this into further practice.

Next dates and further information

Further hackathons are planned for August & September 2026 on the focus topic of AI agents Frankfurt, Munich and Vienna as well as in September in Zurich.

It is advisable to contact the responsible SAP team early on, as preparation can take some time.

The podcast describes the background to the hackathon format here Apple / Spotifyexperiences from the perspective of BVB participants here as Podcast on Apple or Spotify.