Use case · Dashboards without the export
On-premises analytics with governed answers
Dashboards, paginated reports and a governed query layer that run inside your network against your own warehouse — where every number on a screen can say which metric and which query produced it.
- Runs
- On-prem behind SSO, or desktop
- Every figure
- Traces to metric → model → query
- Output
- Dashboards + paginated reports
- Status
- Early access — design partners
Who it is for
For analytics leads who have sat through the meeting where three dashboards disagree about revenue, and whose warehouse is not allowed to be copied into a hosted BI tool.
What this looks like from the inside
Three dashboards disagree about revenue, and the meeting to settle it costs more than the dashboards did. The hosted BI tool holds a copy of the warehouse, the definitions live in a dozen saved queries, and the paginated report the auditor wants is a screenshot.
Drawn, not described
The one thing the prose cannot do: show the moving parts and where the boundary is.
dashboard · finance overview
Net revenue, Q3
- 01 · metric
net_revenue = gross_revenue − refunds − credits
defined once · owner: finance
- 02 · model
fct_orders × dim_customer × dim_date
conformed dimensions from Catalog
- 03 · query
SELECT sum(net) … WHERE fiscal_quarter = 'Q3' AND region IN (:allowed)
row-level security applied at the endpoint
Two charts that show revenue show the same revenue — there is one definition to resolve to.
Step by step
In the order it actually happens.
- 01
Define each metric once
Revenue, active customers, churn — defined in the governed semantic model, beside the conformed dimensions they aggregate over. Not in a saved query, not in a chart's settings.
- 02
Build dashboards against the model, not the tables
Every chart resolves through a metric. Two charts that show revenue show the same revenue, because there is only one definition to resolve to.
- 03
Ask any number where it came from
A figure on a dashboard can show its metric, the model it resolved against and the query it ran. A disagreement becomes a modelling bug with a location, not a meeting.
- 04
Schedule the paginated reports
Statements, filings and operational documents come out of the same model as pixel-accurate paginated reports, delivered on a schedule to the people who need them.
- 05
Let assistants in through the governed door
An AI assistant queries the governed MCP server — metrics and models, not raw tables — and sees only what the asking user is permitted to see.
What actually changes
| Dimension | Today | With Revoluzen |
|---|---|---|
| Where 'revenue' is defined | In a dozen saved queries, each slightly different. | Once, in the semantic model, reused by every dashboard and report. |
| Where the warehouse is | Copied into a hosted BI tool so the charts can render. | Where it was. Studio runs inside the network, behind your SSO. |
| What the auditor gets | A screenshot. | A paginated report whose figures trace to their metric, model and query. |
Which product does what
Each product is adoptable on its own; together they share one catalog, one permission model and one run history.
Revoluzen Studio
Dashboards and scheduled paginated reports from one semantic model; every figure traces to its metric, model and query; on-premises with SSO or as a desktop app.
Product pageRevoluzen Catalog
Supplies the conformed dimensions and the governed SQL endpoint the model resolves against.
Product pageWhere it runs
On an internal server behind your identity provider for a team, or as a desktop application for a single analyst. The warehouse is never opened to the internet.
What you end up with
- 01One metric definition reused by every dashboard and report
- 02Disagreements between reports become a modelling bug, not a meeting
- 03Paginated output for statements, filings and operational documents
- 04A governed MCP server, so AI assistants query the model rather than raw tables
The honest limits
You would find these out eventually. Finding them out now is cheaper for both of us.
Revoluzen Studio is in early access; expect to work with us on your deployment rather than download it and be finished.
It is reporting and dashboarding over a governed model, not a spreadsheet replacement or a data-science notebook.
The visual library is deliberately focused on charts that answer business questions clearly, not on breadth.
Asked about this scenario
Can Revoluzen Studio run entirely on-premises?
Yes. Revoluzen Studio runs on an internal server behind your identity provider, or as a desktop application, against your own warehouse. No data is exported to a hosted BI service.
What is a governed answer in Revoluzen Studio?
A governed answer is a figure that resolves through a defined metric on a governed semantic model, so Revoluzen Studio can show which metric, model and query produced any number on a dashboard or report.
Can an AI assistant use Revoluzen Studio?
Yes, through Revoluzen Studio's governed MCP server, which exposes defined metrics and the semantic model rather than raw tables and respects the permissions of the person asking.
Adjacent scenarios
Medallion, modelled, enforced
A governed lakehouse you operate
Build a bronze–silver–gold lakehouse on open table formats and your own object storage, with contracts enforced between layers, dimensional models as first-class objects, and a SQL endpoint that applies row-level security.
Your model, your keys, your network
AI on your data, on your terms
Use AI to draft pipelines, explain failures and answer questions about your own data — with a model you choose, including one running fully offline, and with every tool scoped to the permissions of the person asking.
Run it yourself, then decide.
Revoluzen Flow is Apache-2.0 and installs in one command. If it fits, we can talk about the parts that are not open source.