At this year's SAP TechEd conference in Berlin, Germany, SAP further ingrained artificial intelligence within its suite of business applications while announcing partnerships to bring together data, AI, and open development principles across a unified operating model for the intelligent enterprise.
Foregrounding the role of developers within the agentic AI evolution, the company unveiled products for AI data management and agent development, as well as a new collaboration with Snowflake, while introducing its first relational foundation model, SAP-RPT-1, specifically designed for tabular business data.
Taken together, the announcements represent SAP's commitment to opening up the SAP product ecosystem to help customers simplify the process of establishing strong foundations for business AI.
- SAP's first relational foundation model, SAP-RPT-1, is described as a new class of AI model built for business users, designed to efficiently and accurate predict the most probable business outcomes (for example, delivery delays or payment risks) rather than predicting probable next words in a sentence.
- SAP and Snowflake announced a partnership to make the latter's AI data platform available as a certified solution extension for companies leveraging SAP Business Data Cloud; simplifying data sharing between the platforms, the partnership will make it easier for customers to connect semantically rich SAP business data with Snowflake's AI agent development capabilities.
- SAP BDC is also gaining a new data product studio capability that makes it easier for customers to turn raw data into data products that support analytics, AI, and application development.
- SAP Build will enable developers to leverage agentic development tools like Cursor, Claude Code, Cline, and Windsurf with SAP development frameworks, using new SAP Build local Model Context Protocol Servers.
- Visual Studio Code users can access SAP Build capabilities directly in their development environment with a new extension that will later be made available on Open VSX Registry for other development environments.
- Joule Studio is gaining new agent building capabilities, allowing users to extend SAP's pre-built agents while building their own.
- Knowledge graph engine capabilities for SAP HANA Cloud will map relationships across SAP database tables, columns, and data models, improving visibility of semantics and business context across systems.
- New Joule agents built for technical users will empower new role-based Joule assistants, providing more specified agents for assistants to coordinate across workflows, departments, and applications (for example, a business process analysis agent to identify process inefficiencies and drive process awareness).
- SAP pledged to equip 12 million people with AI skills by 2030 through an expansion of its Coursera partnership and the launch of new certifications for SAP BTP, generative AI, and SAP Build development.
Introducing SAP-RPT-1
One major SAP TechEd announcement involved SAP-RPT-1, described as the first relational foundation model for enterprise data.
As opposed to large language models pre-trained on unstructured text, SAP-RPT-1 is built specifically for structured business data, including tabular data.
A new transformer-based model embedded natively in SAP HANA Cloud, this open-source model aims to deliver predictions without task-specific pre-training; this "in-context learning," according to SAP, means that the model can deliver fast and accurate predictions around scenarios suchas delivery delays, payment risks, or sales order completion across business functions such as finance, supply chain, and HR — simply by evaluating available relational and structured data and past examples.
SAP is launching a web-based testing playground environment for the new model, available at no cost, to encourage customers to test model performance and learning capabilities using sample datasets or their own business data.
SAP-RPT-1 will be generally available in the fourth quarter of 2025 and delivered via AI Foundation.
SAP Snowflake: Architecting the Business Data Fabric
As SAP seeks to realize its vision of a business data fabric for the enterprise, simplifying customers' data landscapes while preserving critical business context for data to accelerate AI projects, the company has focused on enhancing the capabilities of SAP Business Data Cloud.
SAP's recent progress in this area included the announcement of SAP Business Data Cloud Connect earlier this fall. Enabling bidirectional, zero-copy sharing of data products between SAP BDC and partner systems, this service was first announced for Databricks (already generally available on all hyperscalers) and Google Cloud's Google BigQuery (currently in public preview). SAP expects future such partnerships with more large-scale data sources, such as Microsoft Azure and Amazon Web Services.
Historically, extracting data from SAP applications has required complex replication processes that made it difficult for teams to retain business context, at least without costly reconciliation procedures that involve rebuilding lost metadata through external repositories. Zero-copy data sharing solves this issue by enabling real-time access to data across systems without the need to move or copy it, hence the term "zero-copy."
SAP Snowflake, a solution extension for SAP BDC, will bring Snowflake's data and AI capabilities directly to SAP BDC customers, giving them greater flexibility in selecting the right compute and storage for every data and AI workload. Extending BDC with Snowflake's AI, analytics, data engineering, marketplace, and collaboration capabilities, the partnership will position SAP Snowflake as a cloud-scale compute and storage option for customers. With bidirectional, zero-copy data access, data and AI teams can work in real time with semantically rich SAP data products, within a unified governance framework.
As a result, "customers will be able to harmonize SAP and non-SAP data while optimizing total cost of ownership across workloads and build agents and AI applications in SAP Snowflake fueled by trusted SAP data products," according to the announcement.
Snowflake will support bidirectional zero-copy data sharing between its platform and SAP Business Data Cloud, ensuring semantically rich data can be accessed without replication between SAP and Snowflake's cloud-based enterprise data warehouse platform via access to the company’s Horizon catalog. While SAP BDC Connect is built on the open Delta sharing protocol, it will support additional data formats over time and integrate other technologies to ensure partner platforms can enroll in it efficiently, SAP said.
“By tightly integrating SAP and Snowflake, we’re making it simple for enterprises to connect their critical business data with its rich context in SAP with the power of seamless AI app and data agent development at scale in Snowflake," added Christian Kleinerman, EVP of Product at Snowflake, in a statement.
Connecting the Data Product Economy
Data engineering teams, meanwhile, stand to benefit from data product studio capabilities in SAP BDC, allowing users to create, model, and manage their own data products in a single workspace that makes use of both visual tools and SQL-based transformations.
This visual workplace will help data engineers cut down on data management and collection by centralizing the vantage point from whence they can visualize and control data sources into a catalog.
Elsewhere, SAP detailed an expanded Joule Studio that can now support AI-assisted agent design, system-triggered agents, extensibility of SAP-delivered Joule Agents, centralized enterprise-grade agent monitoring, support for Agent-to-Agent (A2A) protocol, and support for Model Context Protocol (MCP).
MCP support for SAP HANA Cloud is now available, giving AI agents greater access to SAP data in all applications. Additionally, with the introduction of SAP HANA Cloud knowledge graph engine capabilities, SAP HANA Cloud metadata can be served up in automatically generated knowledge graphs that communicate data and context together, making it easier for Joule agents to access the database's relational, graph, and vector data, grounding their reasoning in enterprise semantics.
Similarly, SAP Build is expanding to support AI-based ABAP development, custom code migration tools, and integration with SAP BDC. Via new local MCP servers for SAP Build, developers will be able to use their preferred agentic development tools and code assistants like Cursor, Windsurf, Claude Code, Cline, and OpenAI Codex while maintaining clean-core alignment and governance.
Finally, SAP used TechEd to publicly pledge to equip 12 million people with AI skills by 2030 through an expansion of its partnership with the online learning platform Coursera and the launch of new certifications for SAP BTP, generative AI, and SAP Build development. SAP's emphasis will be on hands-on training and certification programs that integrate practical AI-ready tools.