A new platform should not require giving up control of the serving location.
The budget needs defined capacity and predictable expansion costs.
Your serving location is an explicit choice. That's the design.
Alice reads your sources and writes open files
Database and object-store connectors read the sources you authorize. Uploaded CSV files are stored privately for scheduled processing. Alice writes open-format serving files to your Azure storage, OneLake, or a single-tenant managed container.
Power BI reads those files directly
Alice is not in the dashboard viewing path. It stores encrypted delegated OAuth or service-principal credentials to create and refresh Power BI items; Power BI reads the last delivered serving files independently.
Nothing to install
No on-premises gateway to patch and babysit. Alice fetches from your sources; Power BI reads your storage. There is no agent sitting on a server in your closet.
Last delivery remains available
Dashboards can keep reading the last delivered files if Alice is temporarily unavailable because Alice is not in the dashboard viewing path.
A documented exit path
Files in your storage remain there. Managed-container files are available in open formats during the 30-day post-termination export window, then deleted.
The security and ownership story is not a footnote here. It is the product.
Built for the review your procurement team will run.
The facts a security review actually asks about, stated plainly. No badges we haven't earned, no numbers we can't stand behind.
What the CFO sees
Ordinary run volume does not create a usage-based Alice charge. If you need more than the included capacity, fixed-price pipeline and AI add-ons are available.
If you have a data engineering team and petabytes, buy a warehouse. Alice is not that. Alice is for the Microsoft shop where reporting lands on IT and the budget meeting is already hard enough.