SAP and Databricks extend partnership to bring more AI tools to business data

SAP and Databricks are extending their partnership to bring AI assistance, data engineering, governance and application databases into SAP’s business-data platform. Detailed in an October 30 announcement tied to SAP TechEd, the plan targets a practical obstacle: putting company data to work without maintaining custom extraction pipelines or losing the business meaning attached to it.

The companies introduced SAP Databricks last year, embedding Databricks’ data and AI platform inside SAP Business Data Cloud. The next phase would add Genie, Lakeflow, Unity Gateway and Lakebase to that environment, alongside a planned connection between Databricks’ Genie assistant and SAP’s Joule.

Keeping the business meaning attached

At the center of the data-access plan is SAP Application Data Foundation. The companies describe it as a cloud-native route to SAP application data from cloud and on-premises systems, usable in SAP Business Data Cloud engines, including SAP Databricks. Databricks says the approach is intended to reduce ownership costs and shorten the work needed to get useful analytics and AI running.

The proposed mechanism combines shared SAP data products with SAP Business Data Cloud Connect for Databricks. Databricks says this provides zero-copy access—sharing data without creating an additional copy—and avoids extra extraction pipelines. It also aims to retain SAP’s business semantics: the definitions and context that let models and analytics interpret records as part of a business process, rather than as isolated data.

Databricks illustrates the approach with a hypothetical manufacturer running finance and supply-chain software on-premises while preparing a cloud migration. Its engineers could combine years of SAP order history with outside data to build demand forecasts, then retain that work during the move. The company says customers can work inside SAP Business Data Cloud or bring governed SAP data into an existing Databricks environment.

Four tools, different jobs

The planned additions span several stages of working with enterprise data, from preparing it to building applications that use it. Databricks assigns each tool a different role:

• Genie is the AI coworker intended to give teams access to governed data with enterprise context. Its role is to make the meaning held in SAP data available to employees and agents.
• Lakeflow handles data ingestion, transformation and orchestration: bringing data in, reshaping it and coordinating the steps. It is intended to combine SAP records with sources such as connected devices and customer-management systems.
• Unity Gateway supplies governance and cost controls across AI agents, tools and models. Databricks positions it as the layer for applying guardrails and spending controls as customers extend AI use across SAP data.
• Lakebase is a serverless Postgres database intended to support operational applications and AI agents. In this partnership, its proposed role is to help teams build those applications on SAP and other enterprise data.

The planned two-way integration between Genie and Joule would bring Genie’s answers and actions into SAP experiences. That connection is the announcement’s explicit bridge between Databricks’ AI coworker and SAP’s own assistant, rather than simply another way to move data.

A customer example from the existing platform

The announcement cites chassis-components manufacturer SAF-Holland as an example of the existing SAP Business Data Cloud and SAP Databricks combination. It says the manufacturer unifies more than 100 million customer, delivery and material records, then uses machine learning to predict the reliability of customer delivery schedules. The company-reported result is analyses generated five times faster than traditional reporting approaches—not a performance test of the newly planned integrations.