We have stood up an Enterprise Data and Analytics System (DAS). The tech platform is Microsoft Azure, Databricks and Power BI. We are leveraging the DAS to integrate data for Analytics use cases. We are interested in leading practices and what others do to provision data for application integration. Should our Analytics Platform be leveraged for application integration, or should we have a separate Operations Data Hub, or should reference data for application integration be sourced from the Application System of Record and interfaced directly to the other application? We are interested in the pros and cons of alternatives and industry leading practices.
Sort by:
no titlea year ago
Hello Hasan, thank you very much for your recommendation; this is what we are already doing for Analytics use cases. The question really is whether or not we should use this Analytics platform to provision data for operational application Integration (as well as analytics). For example if we've loaded HR Org data into our Analytics platform for reporting and there is a new (non-Analytics) requirement for that data to be provisioned to a different application (than the System of Record application) for operational purposes - should the data be provisioned from the Analytics System? Thank you very much.
My Recommendation would be as follows:
1-Centralize all data in a data analytics layer (example: Azure Data lake) sourced from the Application System of Records.
2-Build your reporting layer based on the data analytics layer. Have Azure Databricks/Power BI connect to the data lake for the power users within the organization.
3-Publish Power BI Dashboards to end users.
PRO:
-Will not impact/conflict with the back end application scheduled processes and performance.
CON:
-Data lake may not have the real time transactional data.
-Some form of Data Replication services will need to be stood up to feed the data lake on a schedule.