What Business Teams Can Do with Supper and Databricks

Companies often invest in Databricks to bring data engineering, analytics, and AI work into one governed environment. Yet employees outside the data team may still depend on prepared dashboards or analyst requests whenever they need to investigate a new question.

Supper gives those employees a direct way to work with the data their company maintains in Databricks. Users can ask questions in business language, examine the results, and continue the analysis without knowing how the lakehouse is organized. Databricks continues to store, process, and govern the underlying data.

This helps companies get more analytical use from their existing Databricks investment. Supper connects to the Databricks SQL warehouse, uses approved Unity Catalog context, and adds the business logic required for everyday analysis across the company.

Use Supper with the semantic work already in Databricks

Databricks customers may already maintain approved definitions and access rules through Unity Catalog. Metric views provide centrally managed definitions for business metrics, while Unity Catalog organizes the data objects and permissions associated with them.

Supper can ingest supported Unity Catalog context as part of its own semantic model. When the imported definitions are complete and approved, Supper can treat them as canonical. When catalog coverage is still developing, Supper can use that information as input and add context where an analysis requires it.

This preserves work the data team has already completed. Supper can then maintain company-specific Terms, known data exceptions, and instructions for recurring analyses. The same context is applied across users and questions instead of being reconstructed in each conversation.

Databricks also provides native business-facing AI through Genie One and Genie Agents. Supper can be used alongside these capabilities when a company wants a dedicated analytics product with its own semantic model, workflows, and connections to additional business systems.

Give business teams a direct way to investigate questions

Most employees do not think about catalogs or SQL warehouses when they ask a business question. They want to understand why a metric changed, which accounts need attention, or whether the company is performing as expected.

Supper translates those questions into analytical work over Databricks. It uses the company's semantic model to interpret the request, queries the relevant warehouse data, and can use Python when further calculations are needed. The user receives an answer that can be explored through follow-up questions.

Results can also be turned into charts or added to Supper Spaces for continued reporting. The employee can move from an initial question into deeper analysis without exporting the result to another application or asking an analyst to rebuild the work.

This expands access to Databricks data while keeping approved definitions involved in the analysis. The data team can support more users without becoming the required intermediary for every new question.

Analyze Databricks alongside operational systems

Some of the information needed for an analysis may already be available in Databricks. Other details may still live in a CRM, finance application, or customer-support system.

Consider a question such as: "Why did gross margin decline last month?" Revenue and product-usage history may be available in Databricks, while current discounting or other operating details remain in an application used by the business. Supper can work with the relevant sources together and apply the same definition of gross margin throughout the analysis.

Databricks provides its own routes for working with external data, including Lakehouse Federation and Lakeflow Connect. Companies can continue developing that long-term architecture while using Supper to make selected operational sources available to its analytics agents.

The business user does not need to know where each field is stored. Supper determines which available sources are relevant and works with them according to the user's permissions.

Save recurring analytical work as Supper Skills

Many business questions are part of a recurring process. A finance team may review margin every month using the same calculations and investigative steps. Reconstructing that process manually can lead to different methods and inconsistent results.

Supper Skills capture the instructions and analytical logic behind recurring work. Once the margin analysis has been reviewed, it can be saved as a Skill that other people or agents can run again.

The Skill can specify how the analysis should be conducted, which definitions should be applied, and what the finished result should contain. It continues to use Databricks as the source for the relevant lakehouse data.

This gives companies a way to preserve the analytical process used by experienced team members. The work becomes available to other authorized users without requiring them to reproduce every decision behind it.

Control how people access Databricks through Supper

Supper connects to the Databricks SQL warehouse through a customer-provisioned system account. Unity Catalog enforces the warehouse privileges and applicable policies assigned to that account, establishing the boundary of Databricks data available to Supper.

Supper can use supported Unity Catalog entitlement information as an input to its own permissions model. Because warehouse queries run through the system account, Supper does not automatically assume the individual Databricks identity of each employee. The imported permissions and Supper's application-level controls therefore remain distinct parts of the access model.

Administrators can apply Supper's user-level allow and deny rules within the access granted to the connection. Those rules can also cover information from SaaS applications connected through Supper.

This allows administrators to control the full set of data available to each Supper user. The Admin Control Center also shows how people use that access, including the questions they ask and the areas where the semantic model needs more context.

Get started quickly and refine the model through use

Connecting Supper generally requires about five minutes of customer work to provision the Databricks system account. Supper then inspects the permitted environment, incorporates supported Unity Catalog context, builds its initial semantic model, and creates an evaluation corpus for testing.

Users can typically begin asking questions within one or two days. The exact timing depends on the size and condition of the Databricks environment.

As employees begin using Supper, their questions reveal missing definitions or ambiguous business logic. Supper can identify many of these issues automatically. Decisions that depend on company-specific judgment remain with the customer.

During the first month, this review usually requires anywhere from a few minutes to a few hours. The model develops around the questions employees genuinely ask instead of requiring the data team to anticipate every possible use case before launch.

With Supper connected, more employees can use the data and definitions their company already maintains in Databricks. Databricks remains the governed data and AI platform, while Supper gives business teams a dedicated way to turn that data into answers and repeatable analytical work.

Frequently asked questions

How does Supper work with Databricks?

Supper connects directly to a Databricks SQL warehouse through a customer-provisioned system account. Databricks continues to store and process the data, while Supper provides its own semantic model and business-facing analytics experience.

Can Supper use existing Unity Catalog definitions?

Yes. Supper can ingest supported Unity Catalog context and incorporate it into its semantic model. The imported definitions can be treated as approved facts or as inputs that Supper supplements with additional business context.

Does Supper replace Genie One or Genie Agents?

No. Genie One and Genie Agents provide native conversational analytics within Databricks. Supper provides an additional analytics product with its own Business Logic, reusable Skills, dashboards, administrative controls, and connections to SaaS data outside Databricks.

Does Supper move Databricks data into another warehouse?

Supper queries Databricks directly for lakehouse analysis. Companies can keep their existing Databricks architecture and do not need to duplicate the warehouse inside Supper.

Can Supper analyze Databricks and SaaS data together?

Yes. Supper can work with Databricks alongside connected SaaS applications when an analysis requires information from both. Supper applies the relevant business context and user permissions throughout the analysis.

How are permissions handled?

Unity Catalog determines what the Supper system account can query. Supper then applies user-level entitlements that control which parts of the available Databricks and SaaS data each employee can use.

How long does setup take?

Provisioning the Databricks connection generally takes about five minutes of customer work. Supper can usually prepare the initial semantic model and evaluation corpus within one or two days, after which users can begin asking questions.