Release date:
This release extends Priority Monitoring to Power BI Semantic Models, introduces AI-powered regression detection for Tableau testing, and adds Runner-based execution and Column Statistics to the Power BI pipeline API.
As always, you can find important links just below.
📖Full documentation 💡Knowledge base 📞Customer support
Priority Monitoring, previously available for Power BI Reports, now extends to Semantic Models. You can identify which semantic models need attention today, and which are likely to need it in the next seven days, ranked by the same dynamic priority score, now calculated from capacity unit consumption weighted against usage.
Each semantic model entry surfaces key metrics like capacity unit usage and the size of its largest table, along with trend movement so you can spot models heading towards performance issues before they impact reports. You can now also sort the priority table by today's score or tomorrow's forecasted score, and customise the priority threshold to match your monitoring needs.
Want to see it in action? Try it here.
Capacity Planning now includes a Daily Report Openings chart, tracking how often a report has been opened each day, with a forecast for the days ahead.
A clear marker separates past activity from today and the forecasted trend, so you can spot usage patterns and anticipate rising demand before it turns into a capacity or performance issue. Hover over any point for a detailed breakdown of that day's openings.
The configurable folder creation option, introduced in 2026.7.1 for the Power BI Deploy Content step, is now also available for Power BI Deploy Content (Cross Environment), completing support for this feature across both deployment steps.
Administrators can control whether missing folders are created in the target workspace, disabled to fail deployments when required folders are missing, or enabled to create and report them in execution results.
API created and API triggered Power BI pipelines can now execute using a configured Runner's privileges, instead of the credentials of the person scheduling the pipeline. Pipeline ownership stays with the creator, but execution runs under the Runner's dedicated identity. This is available in the API now, with UI support coming in a later release.
API created pipelines for Power BI can now include Column Statistics, supporting both Simple and DAX modes. This brings deeper data quality insights directly into your workflows, with no manual effort required.
Wiiisdom Cloud Platform now offers AI-powered regression detection for Tableau, going beyond traditional pixel comparison. Instead of highlighting broad areas of a dashboard that may have changed, each regression is now broken down into distinct zones, with a plain-language explanation of what changed in that zone, whether it's a title, a filter selection, a KPI value, or a chart's data and layout.
This makes it far easier to pinpoint the actual source of a regression and troubleshoot with confidence.
Customers without access to AI-powered features will continue to see results as pixel-based comparison, as shown above.
For customers with access to AI-powered features, regression results now appear in this new zone-based format, as shown above.
Each detected zone is listed individually with a plain-language description of exactly what changed, whether it's a title update, a filter change, or a shift in chart data, so you can go straight to the source of a regression without having to hunt for it visually.
SQL Data Validation reports now let you explore a greater volume of the tested data directly within the pipeline execution report, making it easier to troubleshoot mismatches and validate data quality at scale.
▪ BI Platform actions now generate audit events
Enabling or disabling Predictive Scoring, and updating a BI Platform, now log audit events.
Predictive Scoring events record the selected Runner, and BI Platform update events record the number of configured Runners.
▪ Admin actions now attributed to the individual user
Administration audit events are now attributed to the specific admin who performed the action, rather than a generic system actor, improving traceability for support and security investigations.
