SAP announced new generative AI and data governance features for SAP Datasphere and SAP Analytics Cloud, enhancing enterprise-wide access to data and insights as part of its strategy to establish a comprehensive business data fabric for customers.

The innovations, unveiled March 6 during the SAP Data Unleashed virtual event, are meant to provide customers with a unified data view and simplify their landscapes while retaining the business context and logic required to effectively leverage generative AI.

Juergen Mueller, SAP Chief Technology Officer and member of the Executive Board of SAP SE, described the data, planning, and generative AI capabilities as a “quantum leap” in the company’s ability to help customers drive intelligent business transformations with data, noting during the event that “implementing a holistic data and analytics strategy” has been a key priority and challenge for C-level executives, including CIOs, in the SAP ecosystem.

The announcements included:

  • SAP Datasphere will allow organizations to better represent their real-world use of data through a knowledge graph, providing more complete context from disparate data sources to large language models (LLMs) while inhibiting AI-generated model hallucinations.
  • SAP HANA Cloud vector capabilities, to be introduced in the next release of SAP HANA Cloud, will support generative AI outputs with business context and further empower customers to build intelligent data applications.
  • Improved integration between SAP Datasphere and SAP Analytics Cloud will establish a single data management system with advanced analytics, to support customers’ data preparation, modeling, and planning.
  • Joule, SAP’s generative-AI copilot, will be made available in SAP Analytics Cloud, to automate the creation and development of reports and dashboards.
  • A “compass” functionality in SAP Analytics Cloud will enable business users to run data-driven simulations using a chat interface to evaluate predictive outcomes and continually adjust controllable variables, optimizing customers’ scenario modeling capabilities.
  • SAP will expand its partnership with Collibra to integrate its AI Governance platform with SAP data assets, enabling data governance for non-SAP data assets and more fully supporting customers’ heterogenous landscapes.

Weaving a Business Data Fabric

SAP Datasphere, launched a year ago, is intended by SAP to serve as a “comprehensive data service” for customers, helping them to unify their enterprise data, break down data silos, and centralize data management. Built on SAP Business Technology Platform and first introduced as the cloud-based next generation of SAP Business Warehouse, SAP Datasphere facilitates data integration, cataloging, semantic modeling, data warehousing, and virtualizing workloads across both SAP and non-SAP data.

Core to SAP Datasphere is the concept of an overarching “business data fabric,” a data management architecture that establishes an integrated, semantically rich data layer over customers’ existing data landscapes. The architecture provides seamless, scalable access to any data across an enterprise without duplication, while keeping business context and logic intact.

With these new features, SAP is updating its data strategy for the business-AI era and acknowledging that organizations cannot effectively leverage generative-AI innovations without trust and transparency surrounding the process of data capture and the quality of their data.

By bringing data governance, data modeling, and generative AI features to SAP Datasphere, the company is moving to prioritize data quality and access while ensuring non-SAP and unstructured data (such as text, images, and audio) can also be effectively leveraged in AI-based scenarios.

“Capturing data to make better decisions is an enterprise technology imperative that’s become increasingly critical as AI—which relies on quality data—revolutionizes every aspect of business,” said Mueller, in an official statement.

During the Data Unleashed event, he positioned SAP Datasphere as the solution to solve this challenge and urged customers to “put the right data architecture in place that will enable you to manage big data sets without data redundancy, gain valuable insights, and make intelligent decisions—all at scale, without rebuilding the semantics you already have in your SAP system.”

SAP Datasphere Knowledge Graph

SAP Datasphere’s knowledge graph will take existing data onboarded, transformed, and integrated in SAP Datasphere and automatically create an ontology showing relationships and connections between the data, including business context from other SAP applications like SAP S/4HANA.

Via an ontology editor, customers can customize the ontology to fit their specific business needs; then, existing data within SAP Datasphere will be combined with the ontology to create a knowledge graph, presenting and processing the data “as a semantic web of relationships,” as per SAP.

“Knowledge graphs are powerful tools for capturing the rich and complicated relationships that exist across your enterprise by combining them into your business data fabric,” Mueller said during SAP Data Unleashed. Combining business data with relational metadata, he added, is another benefit; with knowledge graphs, business users can understand more completely how the data in their landscapes is being used, improving their ability to uncover hidden connections, insights, and usage patterns between applications.

Additionally, knowledge graphs are equipped to answer complex, open-ended questions—posed by business users through Joule—and can provide context to LLMs, reducing AI-generated “hallucinations” in data. “Knowledge graphs are crucial for business AI,” said Mueller. “Latest research shows that knowledge graphs are more accurate than vector engines when dealing with structured data in the context of LLMs, and the flexible structure is ideal for answering various questions, especially open-ended questions.”

The functionality is in preview with select customers. The announcement builds on last year’s announcement of an Analytical Model in SAP Datasphere for multidimensional and semantically rich analytical modeling, allowing customers to better answer business questions.

SAP HANA Cloud Vector Engine

With SAP HANA Cloud, its multi-model database management system, SAP already provides developers with the ability to build and deploy intelligent data applications. A single, integrated data platform that can store, access, and process different types of data, SAP HANA Cloud serves as the database management foundation of SAP Business Technology Platform.

Vector capabilities in SAP HANA Cloud will make it easier for AI models to manage unstructured data and enhance interactions between LLMs and organizations’ mission-critical data. Customers will be able to supply business context to ensure they are present in generative AI outputs.

Retrieving data and context in the native language of LLMs, vectors (long arrays of numbers that give value to a large number of attributes or dimensions) will allow SAP HANA Cloud users to gain more context on their business data. With this functionality, the SAP HANA Cloud database can be leveraged to continuously train LLMs underneath organizations’ generative AI applications with up-to-date, relevant business information. According to SAP, this will simplify interactions with LLMs and encourage developers to build generative AI into their own applications.

With new vector database capabilities enhancing the value that customers can gain from its existing multi-model capabilities, the SAP HANA Cloud vector engine will additionally allow customers to store and compare vectors in SQL, simplifying data architecture and security. Customers will be able to combine spatial, graph, JSON, and relational data with vector queries. Additionally, vector databases can be used for retrieval-augmented generation (RAG), recommendations, classifications, and clustering, according to SAP.

This announcement builds upon SAP’s previously announced vector data store for SAP HANA Cloud. SAP HANA Cloud’s vector engine is in public preview with select customers, with general availability planned for late Q1 2024.

Integrating SAP Datasphere with SAP Analytics Cloud

To establish SAP Datasphere more fully as customers’ platform of choice for integrating planning processes and tools, SAP will more closely integrate SAP Datasphere with SAP Analytics Cloud, including by making it possible to deploy planning models from SAP Analytics Cloud into SAP Datasphere.

Extending SAP Analytics Cloud’s planning functionality via SAP Datasphere will reduce the complexity and risk that customers face in storing multiple copies of data in different systems, bringing plan and actuals data together to enable real-time steering and further establish a business data fabric across the SAP solution suite.

Integrating SAP Analytics Cloud with Generative AI

To accelerate a holistic data and analytics strategy for customers, SAP is integrating SAP Analytics Cloud, its solution for business intelligence and planning, with its generative-AI copilot, Joule, for data discovery, dashboard creation, and planning model maintenance. Joule will leverage SAP HANA Cloud’s vector engine to do so, ensuring that it maintains visibility of business context and logic.

Meanwhile, through the “compass” feature for SAP Analytics Cloud, building upon previous Monte Carlo simulations in existing SAP tools, business users can more efficiently predict or forecast business scenarios using enterprise data, without advanced statistical skills. Business simulations can be performed directly in SAP Analytics Cloud, with scenario modeling capabilities allowing for comparisons between assumptions and improving users’ ability to analyze relationships between data.

A preview of SAP Joule functionality in SAP Analytics Cloud is planned for the end of Q2 2024, while a preview of SAP Analytics Cloud compass is planned for the end of Q3 2024.

Expanding the Open Data Ecosystem

SAP announced it will develop an AI governance solution in collaboration with Collibra, which launched its own AI governance solution last month. By extending its partnership with Collibra (whose SAP and non-SAP metadata management and governance solutions were first integrated with SAP Datasphere at launch) to include AI governance, SAP will enable customers to see what data their various AI models have been trained on, to ensure that data sources remain trusted and accurate.

“We know that SAP data must be open and work with your existing data infrastructures,” said Mueller during SAP Data Unleashed. “That is why we announced the open data ecosystem last year, and we are continuing to deliver new capabilities with our partners.”

Additionally, SAP announced an integration partnership with Confluent. Through this partnership with Confluent for real-time analytics and decision-making tools, customers can stream and connect business-ready SAP data to external data and applications in real time. “Customers can now send the right signals to Kafka-end points with SAP Datasphere data integration capabilities and soon replicate data from Confluent into SAP Datasphere as well,” Mueller said.

For more information on all the announcements, visit SAP.

Like what you’re reading?

Become a member and get access to all ASUG benefits including news, resources, webcasts, chapter events, and much more!

Learn more

Already an ASUG member? Log in