SAP and Databricks Deepen Partnership, Launching Application Data Foundation to Simplify Enterprise AI Access
SAP and Databricks announced an expanded partnership featuring the new 'SAP Application Data Foundation,' enabling zero-copy access to SAP business data within SAP Databricks and integrating AI capabilities like Genie.
German software giant SAP and US data intelligence company Databricks announced a deepening of their partnership at SAP TechEd on October 30, 2026. The two companies introduced the new 'SAP Application Data Foundation,' designed to provide simpler and more performant access to data from SAP applications within SAP Databricks, covering both cloud and on-premises environments. This move addresses a core pain point in enterprise AI adoption: the lack of contextual understanding of actual business processes.
According to the official announcement, over 90% of Fortune 500 companies use SAP systems to manage key business processes. Traditionally, leveraging this data for advanced analytics and AI required building and maintaining custom extraction pipelines, which led to data copying, fragmented governance, and loss of business semantics. The new 'SAP Application Data Foundation' offers a cloud-native way to access this data, allowing direct consumption in the SAP Business Data Cloud engine via zero-copy technology without additional ETL processes.
The update fully integrates several key components of the Databricks platform into SAP Databricks. These include: Genie, an 'AI coworker' providing instant access to governed data and AI; Lakeflow, a unified data engineering solution; Unity Gateway, for multi-AI governance and cost control; and Lakebase, the first serverless Postgres database built for AI apps. Together, these tools enable customers to build models, apps, and agents grounded in SAP's business semantics.
As a case study, SAF-Holland, a global manufacturer of chassis components for trailers, trucks, and buses, uses SAP Business Data Cloud and SAP Databricks to unify more than 100 million records of customer, delivery, and material data. The company applies machine learning to predict the reliability of customer delivery schedules. According to vendor disclosures, SAF-Holland now generates analyses 5x faster than traditional reporting approaches, turning complex SAP data into planning insights in minutes instead of weeks. While these benefits come from vendor-reported customer stories, they illustrate the potential application of this architecture in real-world industrial scenarios.