How to implement transparency in a hybrid landscape with the help of SAP Datasphere and Python?

 

Data models, dimensions, and structures are not the same throughout the landscape. SAP ERPs may have traditional General Ledger, new G/L or S/4HANA simplified structures – which are slowly changing through the versions until all simplifications are fully done. This gives the first challenge.

Another challenge is slowly changing dimensions. The chart of Accounts has small changes, and new profit and cost centers are introduced every now and then. New S/4HANA may require additional dimensions and a new organizational model. During the S/4HANA program, you may not have a single source of truth, instead, you need to maintain master data in several systems and harmonize for consolidation and group reporting.

S/4HANA should follow the clean core principle - it is not the best solution to customize the data. Here SAP Datasphere comes into the picture. It has a space concept and replication flows for data. You can easily replicate data from SAP source systems using real-time delta replication. You can further aggregate the data in Datasphere according to your needs, harmonize, and perform transformations.

Power of Python and Datasphere

In our projects at BearingPoint, we have used Python coding on Databricks, on top of Datasphere. Python is, of course, flexible for data modification purposes. You can read master data and drivers from Datasphere tables, either fixed or dynamic values, and implement quite easily a complex distribution and allocation logic.

Here we can harmonize, transform, distribute, and allocate data in very effective ways, without any special SAP tool competence. There are more Python-capable developers than SAP Datasphere competent in the market currently. Both are needed in the end, of course, but by using different skills we can speed up the development and be much more flexible compared to traditional ETL (extract, transform, load).

An example of components in an SAP hybrid landscape, to be aligned with a customer-specific landscape.

End2End transparency in a hybrid architecture

On SAP hybrid architecture SAP Datasphere can act as a central point between source systems and feed prepared data to several target systems, for example, for planning, statutory, and group reporting.

Hybrid architectures are here to stay. We can use data federation, we can call this a data mesh, but all in all, by using the right toolset and competencies we can be productive. SAP Datasphere should be an integral part of your data mesh architecture but also enable real-time access to your SAP data where you need it. The balancing of the right approaches in an ever-moving enterprise landscape is where BearingPoint can help you with references and practical experience.

Author

Tapani Tuoma
Senior Business Advisor, ERP team
BearingPoint Finland