Request Summary:
We are seeking enhancements to the current user permission model within Microsoft Fabric/Power BI to better support our self-service BI strategy while maintaining governance and performance standards.
Background & Business Need:
As part of our ongoing digital transformation and data democratization efforts, we are encouraging broader self-service analytics across the organization. However, the current permission model in Fabric/Power BI lacks the granularity required to effectively balance empowerment with control.
Specifically, we need to ensure that users can explore and build insights independently, but within defined boundaries that prevent misuse or performance degradation. For example, we want the ability to assign or restrict user permissions based on the type of artifact they are allowed to create. For example:
Allow: Exploration, Dashboard
Restrict: Report, Paginated Reports, Real-time analytics, etc...
Equally, we want flexibility here so that another user group, we may want to allow more, i.e. reports may be allowed.
This would enable us to guide users toward appropriate tools for their skill level and use case, while reserving more complex or resource-intensive capabilities for advanced users or central of exellence teams.
To build on this I would like better Dataset (Semantic Model) Usage Controls:
To prevent performance issues and ensure efficient use of shared datasets, I would like to see:
The ability to limit the amount of rows specific user groups can call - simialar to RLS policy
Controls to restrict the complexity of queries i.e. Memory/CU use by User Group
Expected Benefits:
Improved governance and performance of shared datasets.
Reduced risk of poorly designed reports impacting system stability.
Empowerment of users through self-service, within safe and manageable boundaries.
Better alignment with our data strategy and operational efficiency goals.
