Skip to main content

THE DATA OS · DASHBOARD LIBRARY

Live examples, not static screenshots of 3 year old case studies.

Every dashboard in here is the real report, running in your browser. Filter it, split it, sort it, try to break it. Then notice the thing a screenshot can never do: the tables underneath are a question away.

Why this library exists

Most dashboard galleries are a slideshow. Somebody screenshots a report, crops the numbers out for privacy, and posts it as a template. You get to look at the shape of someone else's answer.

That is the same failure we watch happen inside companies. A report gets built once, screenshotted, and dies on somebody's desktop. An operator put it to us this way: the AI creates an incredible report, then hands you an HTML file. It has nowhere to live, no way to refresh, no way to share.

So these are live. Every dashboard above runs on a governed pipeline, every number on it has a definition written down, and the tables underneath it are queryable in plain English through your organization's MCP. That last part is the point, and it is the part a screenshot cannot fake.

Organized by business model, because that is how you know whether one applies to you. We would rather ship four dashboards with every number defined than forty screenshots with none.

What every dashboard here has in common

Live, not a picture

The embed on each page is the actual report. Change the date range, split the funnel by device, sort the table by revenue. Nothing here is an image of a number.

Every number has a definition

Each dashboard ships with the written definition of every step it counts. Nobody should have to guess whether a lead means a form fill or a qualified opportunity, and on these pages you never do.

Queryable through your MCP

The dashboard is one view of the tables beneath it. Through your org's MCP, anyone on the team can ask those same tables something nobody anticipated and get an answer that uses your definitions.

Lives in a warehouse you own

In a real deployment the data lands in a BigQuery wired to your billing account. Your existing tools become sources instead of homes, so nothing is held hostage when priorities change.

A screenshot can only answer the question it was cropped for.

Every dashboard you have been sent as an image was somebody else's answer to somebody else's question. The version you can filter, split, and query is the one that answers yours.

Questions about the library

Is this real client data?

No. Every dashboard in the library runs on synthetic rows generated for the demo, so nothing here exposes a client. The structure, the metric definitions, and the pipeline behind them are exactly what we deploy.

Can I get one of these built on my own data?

Yes, that is the actual product. The Data OS syncs your sources into a warehouse you own, defines your metrics once in a glossary, and puts living reports on top. The first live dashboard usually lands inside 2 to 4 weeks.

What does it mean that the data is queryable through the MCP?

Your organization gets one MCP endpoint that Claude, Cursor, or any AI client can connect to. Anyone on the team asks a question in plain English, and the answer is built from the same governed tables the dashboard reads, after consulting your metric definitions. No SQL, and no waiting on an analyst.

Are these dashboard templates I can download?

No, and that is deliberate. A template is a layout. What makes a dashboard trustworthy is the pipeline and the definitions underneath it, and neither of those travels in a file. What you can take from these pages is the thinking: which steps are worth counting, and what to call them.

Ready when you are.

Bring the report your team argues about most. Thirty minutes with JJ, no pitch, and we will map what it would take to make it live.

Book a Custom Strategy Call