Power Automate: Process mining - Get support for normalized schema import for data ingestion

Most systems of record have data stored in data structures that allow you to easily extract information about the analyzed entity (case-level data) and historical changes on those entities (event-level data). The attributes are also stored in a normalized form. The single-table approach to event log creation requires you to do inefficient transformations. With this feature, Power Automate Process Mining supports the case-event entity-based data structure plus a normalized form of attributes. This speeds up the ingestion process in both the transformation and import to process model phases and reduces the storage capacity needed.

Business Value

Customers with existing data pipelines or data products in data mesh mostly want to stick to normalized, efficient forms such as star schema. Power Automate Process Mining currently requires you to transform the event log into a single-table form, where attributes with small cardinality are duplicated. Support for star schema reduces storage volumes and speeds up ingestion performance.

Enabled for

Enabled by admins/makers/marketers for end users