Tracking User Interactions with Semantic Models via Audit Logs and APIs

[v-erakesh1]

Hello Team,

We would like to propose an enhancement to enable detailed tracking of semantic model usage. Specifically, we need the ability to capture, for each user:

• When they accessed a semantic model
• What activities or operations they performed

Currently, we have options such as usage metrics for reports and the “View Report” activity, which allow us to retrieve logs for report-level interactions. We would like similar capabilities for semantic models, either through audit logs or APIs, so that organizations can maintain comprehensive tracking and governance of semantic model activities.

[Erik_Eld]

It's already possible to track who and when someone accessing semantic models by using the activityevents api. At least when it has been access in Power BI desktop or Excel. We use it for monitoring the usage of reports specifically developed for Excel consumption. However, the logging is very limited and for example it's not specifically mentioned that Excel accessed the model, only that the User Agent is MSOLAP. When accessed from PBI Desktop, the "App" is filled with "Power BI Desktop", but from Excel it's blank. This is just an example. We would like the logging to be significantly improved so that we better can monitor and govern the usage of semantic models.