Data Collections

A Data Collection is a workspace built around a specific set of data: the sources it comes from, what the terms mean in your business, and the queries your team trusts.

Audience: builders

What makes it more than a connection

Your administrator connects the warehouse or the system of record once. A Data Collection is what you build on top, and three things make it worth building.

Queries run live against the source. Zero-copy: the data is not duplicated into the platform. What you get back is what is in the warehouse right now, and your existing access controls still apply.

It speaks your vocabulary, not your warehouse's. Tables do not speak your business; entities do. You define what a Customer, an Invoice, a Vendor, or Days Payable Outstanding means, and questions get translated into SQL for you. Your team asks in their language rather than in column names.

Answers carry their lineage. Every result records what query ran, who ran it, and what came back. Auditors get the trail rather than a paragraph in a memo.

Trust comes in two layers

Asking a question the collection has never seen works, because it can compose SQL from the catalog. That is what makes it useful immediately.

For the questions that matter most, verified queries pin the SQL your team approved, so everyone gets the same answer to the same question. Use the first layer to explore and the second layer for anything someone will act on.

Set one up

Create a Data Collection→Start here. From clicking + Data Collection to an empty Setup tab ready for sources.Add data sources→Attach the sources IT has connected. A collection can blend several behind one queryable surface.Scope the collection→Choose the tables and columns that are in, and the ones that are out.Build the catalog→The semantic shape: entities, relationships, metrics, and the descriptions that tie them to your tables.

Use it

Verified queries→SQL your team approved and published. Same input, same output, every time.Use a collection→Attach the collection to a Project and start asking questions from chat or from a Workflow.Use it from another AI tool→Every collection publishes an MCP endpoint. Point Claude, Cowork, or any MCP client at it.Dashboard→The collection's landing surface: what it holds, what changed, and what is being asked of it.

Project or Data Collection? If you are describing something your team does, that is a Project. If you are describing data your team owns, that is a Data Collection. Projects attach the collections they need.

Custom views Create a Data Collection