SAP Sets Databricks Alliance, Launches Business Data Cloud

The new data management system incorporating Databricks data engineering and AI capabilities is targeted at improving business access to data for analytical, digital transformation and AI initiatives.

SAP today unveiled SAP Business Data Cloud, the latest generation of the software giant’s data management platform for unifying data, both from SAP applications and third-party systems, for analytical and AI tasks.

In addition to incorporating – and building on – current SAP products including Business Warehouse, Datasphere and Analytics Cloud, Business Data Cloud offers natively embedded data engineering, AI and machine learning technology through a new OEM deal with Databricks.

Business Data Cloud includes new capabilities and pre-built content including packaged data products derived from SAP applications and what the company calls “insight applications” that use that data and AI models connected to real-time data for advanced analytics and planning across lines of business such as finance and human resource management.

The new data platform also boosts the operation of Joule, SAP’s AI copilot that launched in September 2023, for cross-functional workflows and business decision-making.

SAP also debuted a series of ready-to-use Joule agents that work with SAP finance, service and sales applications, and new AI agent builder capabilities that allow customers to extend Joule agents and build and deploy their own AI agents. A key component of SAP’s agent infrastructure is the company’s Knowledge Graph data modeling technology, which SAP debuted one year ago, that semantically understands relationships across data sets and makes it possible to discover insights and patterns across applications and systems.

“AI has the potential to create exponential value for organizations around the world, whether that's in the form of automating tasks that require a significant amount of human engagement today or, in some cases, doing tasks that just couldn't be done or weren't practical to be done by any amount of human engagement in the past. Or also to provide the insights with predictive capabilities with next best action recommendations to allow for better, more efficient decision making,” said Muhammad Alam, SAP head of product engineering, during an online press briefing Wednesday.

“But what our customers are telling us is that they're struggling to implement and realize the value of AI in their organizations,” Alam said. “It's hard to get reliable data, high-quality data sets across the enterprise, to power this AI. Many have done hundreds of [proof of concepts] but have not been able to get them to production effectively. And the ones that have been able to get to production haven't really seen the adoption that they're looking for to be able to realize the value.”

SAP is uniquely positioned to provide these capabilities, Alam said, because of the company’s business applications that create data used for data analytics and AI models, which in turn support business processes and workflows that benefit from AI and analytical results.

SAP, like many IT companies, earlier integrated its software with the Databricks Data Intelligence Platform. But Alam escribed the new alliance as a “truly one-of-kind partnership in the tech space.” During the online press conference Irfan Khan, chief product officer for SAP data and analytics, noted that many SAP and Databricks customers already use both companies’ products.

“There’s all this investment in AI, but ultimately AI is only as good as the data,” said Shanku Niyogi, Databricks vice president of product management, in an interview with CRN. “Everyone wants to get more value out of their data and SAP platforms have some of the most valuable enterprise data.”

“Most companies today already have a data strategy in place,” said Daniel Yu, chief marketing officer for SAP data and analytics, also in an interview with CRN. They have assembled data warehouses and data lakes surrounded by data governance policies, he said, and now they are trying to leverage those systems for digital transformation and AI initiatives.

But Yu said these efforts have created “what we call the data divide, the disconnect between the large investment that [organizations] have made in data platforms and the data that companies need to be able to run their business…this is between IT and the business users.”

Yu said that the SAP Business Data Cloud is designed to bridge this divide with its pre-built data content, embedded Databricks capabilities, and SAP’s existing Datasphere, Analytics Cloud and Business Warehouse systems. He emphasized Business Data Cloud’s ability to incorporate data from SAP and non-SAP systems from both cloud and on-premises sources.

Through the relationship with Databricks, every Business Data Cloud user can easily move SAP data into the Databricks Data Intelligence Platform and have immediate access to its capabilities, Niyogi said, pointing to AI-assisted data science and SQL data warehousing as examples. They also will have access to the entire stack of Mosaic generative AI capabilities that Databricks acquired when it bought startup MosaicML in 2023.

SAP Business Data Cloud actually works by providing access to the serverless Databricks platform using integrated permissions and relies on the Delta Sharing open protocol developed by Databricks that allows organizations to securely share data across different cloud platforms without replicating or copying the data. While Databricks has data integration relations with many other IT vendors, Niyogi said the SAP alliance integrates the two companies’ technology at the business model layer.

“So from an SAP perspective, it almost feels like a native application even though we're hosting that [Databricks] service ourselves,” Niyogi said. “So this is a new level of integration for us, probably one that we will explore with others. But we're excited about the opportunity that this creates for SAP given just the sheer size of the important, mission-critical data that SAP has.”

“Every company on the planet wants to get more value out of their data and greater returns on their AI investments,” said Ali Ghodsi, Databricks co-founder and CEO, in a statement announcing SAP Business Data Cloud. “By joining forces with SAP, we’re helping organizations bring together all their data – regardless of format or where it lives – to govern, analyze and build domain-specific AI applications on the Databricks Data Intelligence Platform.”

For partners, Yu said Business Data Cloud will make it easier for systems integrators and solution providers to build and deliver data solutions to their customers and develop data-centric applications that run on the new system.

Datasphere is SAP’s platform-as-a-service data fabric offering for building data pipelines. When asked during the online press conference about Datasphere’s future, SAP executives acknowledged that Datasphere is “evolving” into Business Data Cloud.

Business Data Cloud will initially be available on the Amazon Web Services platform followed closely by availability on Google Cloud Platform and Microsoft Azure.

This story was updated with additional information from Databricks at 8:20 pm ET Feb. 13, 2024.