With AI and SAP to the intelligent enterprise

The consistent focus of the SAP portfolio on the transformation into an intelligent enterprise makes it possible to operate the core business in a single platform and at the same time to develop and test new services and new lines of business in a kind of innovation environment.

Above all, the intelligent technologies available in the SAP product offering are a key success factor for the use of AI: they improve the user experience, automate processes, and create directly measurable added value through cost reductions.

Worth mentioning here are the Analytics, Data Intelligence and Machine Learning modules of the SAP Leonardo ecosystem in conjunction with the SAP Cloud Platform and Hana.

These form the foundation for the operationalization of AI methods. The openness for the programming languages R and Python also enables SAP to meet the requirements of the data science community.

Scaling crucial for benefit

While some companies have anchored AI in their overall strategy, in practice many AI projects are seen merely as research trials: They serve to elicit the value of AI - and end up already at pilot level.

The decisive factor for the benefit of AI projects is the use case-based approach according to the principle "think big, start small, scale fast": In this process, we methodically develop use cases for the application of AI - for example in the form of design thinking workshops - and determine their added value in proof of values.

Those with the highest potential are implemented productively and ideally rolled out across the Group to achieve the highest possible scaling effect.

This approach works particularly well in the SAP environment, because SAP's Intelligent Suite provides extensive standard content for various company divisions and business areas on which AI methods can be built.

For AI-based SAP applications, the Hana Full Use license is recommended to store results from AI models or to read in external data, such as from social networks.

The Hana Runtime license found at most customers, on the other hand, significantly limits the ability to develop AI applications. Furthermore, the Python interface demanded by the data science community should be mentioned.

Hurdles with SAP

Python is seen as one of the most important programming languages in the AI context. However, not as many integration possibilities exist for it as through the R interface.

However, SAP has taken an important step with the Data Intelligence module: It offers a development environment for data scientists and considerably expands the standard range of functions with the option of using R and Python.

Nevertheless, it will be exciting to see which other Python interfaces SAP will retrofit in other modules.

In our experience, the SAP platform can provide a solid foundation for innovation by dovetailing "run the business" (core business) and "win new business" (new business).

Our recommendation is to check the suitability of AI for the individual company in case of doubt using the use-case-based approach. Currently, only a fraction of use cases result in productive use.

There are many reasons for this, but in particular it is due to a lack of availability of relevant data in sufficient quality. Data management - as the basis of all AI applications - is and remains essential.

The highly integrated SAP system is the most widely accepted gold standard for many processes and thus the right environment for initial applications.