Migrate semantic layer SAP BO to Power BI or Looker

 

Migrate semantic layer SAP BO to Power BI or Looker

The SAP BO semantic layer has been around for almost 30 years   A long time ago, Business Object (later acquired by SAP) created a revolution with the concept of a “semantic layer” to facilitate the understanding and interpretation of data by businesses: the famous Objects, organized in Universes. A very daring first step towards what was not yet called “Self BI”.   The world of Analytics is evolving. The rise of the Cloud and the absence of a very precise roadmap for BO from SAP have led a certain number of companies to consider a migration from tool to tool. A particularly complex migration.  However, BO is a proven and robust technology which - with its (very) long experience of its semantic layer - has been able to organize information in companies.      The idea is therefore to allow teams to quickly start from an optimized semantic layer in the target technology, to be able to independently build the data visualization layer of tomorrow.    Ellipsys worked for a large distribution group on a migration of the semantic layer of an SAP BO platform to Power BI and Looker. Our bias, as usual, was to automate the entire project.

Essentialize the migration of  SAP BO Universes

Detach yourself from the physical data model of the Universes:  Like the reports produced, only a small part of the information exposed in the Universes is actually consumed. {openAudit} ,  our solution,  allows you to identify what is really used by BO users by formally identifying the sources of SAP BO dashboards. This allows us to keep only the essentials and “erase” a significant part of the complexity of SAP BO.

The migration of the SAP BO universe to  Looker

{openAudit}  allows you to transcribe the SAP BO Universes to Looker by following this logic:  Objects are defined in Looker view fields in Explore;  The entire classification and nomenclature of the Objects is reproduced in the target;  If necessary, the “semantic” intelligence is converted into LookML.    Unlike the BO universes, Looker's Explores are mainly based on the notion of star or snowflake models. Our engine therefore makes it possible to break down the complexity of the Universes into several simple models which can then be reconciled at the reporting level.   The particularities specific to the universes such as the notions of analysis context, aggregates, nested objects are included with the native functionalities of Looker and/or with Liquid in LookML

The migration of the SAP BO universe to  Power BI

There isn't really a semantic layer in Power BI. Also, to transcribe the semantic layer from SAP BO to Power BI,  {openAudit}  reconstructs the data models in the Power BI Datasets.   We can think of Datasets as the first layer of data exposure for Power BI users. This layer offers a simple organization of business information by ignoring physical data, like the BO Universes. Another possibility is to organize "cubes" directly in Azure SQL and/or MSSQL to provide further organization of company information.

Conclusion    If, like others, you are planning to decommission SAP BO, the question of the semantic layer will necessarily come up. It will perhaps even be the keystone of your migration if the project is to start again on new bases, by allowing the teams to rebuild the data visualization layer by hand.  We therefore offer to automatically transcribe, and for a fixed price, the SAP BO semantic layer to Power BI or Looker. We are working on other projects.    PS: we also offer  complete "as is" migrations of SAP BO and Power BI / Looker reports  : including layout and intelligence.        Read also :    Moving from Teradata to BigQuery? Automate the process!