AI empowered SAP embedded analytics

SAP INNOVATION OPPORTUNITY – we are looking for co-designers and co-investors to build the first MVP of a new exciting AI solution for SAP embedded reporting. 

If we have a closer look on the CDS views, we can see that they are built as a dependency with other CDS views and with transparent SAP tables. When exposing a CDS, we can get its Metadata in xml format, which give us the list of all the objects we can use in the embedded reports. These information need to be accessible by the target AI solution – the exact way of this access is under evaluation. 

STEP 1: The business is requesting a new analysis on Sales amounts and quantities. To prepare the definition of the target Sales report, the AI can already provide a support to guide the users by the expression of the business requirements. In below example, the AI will provide the SAP CDS based data catalog to the users. If the business use these terms to the report definition, the quality of the results in the following steps will be significantly increased.

STEP 2: The business user defined his Sales Report 1 and would like to know which CDS reports could provide the data from SAP S/4. The AI solution would generate a proposition with the corresponding structures and indicating the missing information from the standard CDS views. The IT teams can provide some customer parameters, example restricting the CDS views to be analyzed, routing to standard and/or customer CDS views.

STEP 3: The SAP or BI team can request the AI solution to provide a proposition for the enhancement of the proposed CDS view. The generated code can be replicated to S/4 via AbapGit or manually. Client parameters  can influence the generated code (naming standards, custom BADI to use, etc.) 

STEP 4: Before handing over the proposed CDS report to the business, the IT team can request the AI solution to generate the documentation of the new SAP report and to prepare test scripts (can be automated or manual test scripts). The test scripts can use sample data provided for the testing, which are ideally available in a central Test data DB. The requestor business user validates the new CDS based report.