From Software Vendor to Business AI Company
SAP’s leadership is openly asking whether the company will remain a traditional software vendor at all. In recent keynotes, the answer came back from its own AI: SAP is repositioning itself as a business AI company. Instead of treating AI as a thin feature layer on top of ERP, SAP is introducing an autonomous enterprise platform composed of the SAP Business AI Platform and the SAP Autonomous Suite. The vision is that core applications for finance, supply chain, procurement and HR no longer simply record transactions; they host AI agents that reason, recommend and act inside end‑to‑end workflows. This shifts SAP’s value proposition from selling standalone applications to owning the control layer that orchestrates data, process logic and execution across landscapes. For customers, the promise is enterprise automation that is deeply grounded in business context rather than generic copilots bolted onto fragmented systems.
The Technical Architecture: Data, Knowledge Graphs and AI Agents
At the heart of SAP Business AI is a governed data and context layer designed to keep AI agents aligned with real-world operations. SAP argues that no AI agent can compensate for a bad data landscape, so it is unifying SAP Business Technology Platform, SAP Business Data Cloud and SAP Business AI into a single environment. A central Knowledge Graph maps business entities, processes and relationships across a customer’s estate, giving AI agents a consistent view of master data, policies and transactions. Hundreds of managed data products are available out of the box, with a data-product generation agent to model new ones faster. This foundation lets AI agents in the autonomous enterprise platform reason across information from any source and environment, not just SAP systems. Instead of merely improving analytics, the data layer becomes the substrate for AI agents ERP deployments that execute governed, auditable business actions at scale.
Autonomous Suite in the Wild: Finance, Procurement and Physical AI
SAP’s Autonomous Suite extends this architecture into concrete AI agents across business functions. The suite is set to deploy more than 50 domain-specific Joule Assistants, complemented by over 200 specialised agents for narrower tasks in finance, supply chain, procurement, human capital management and customer experience. Examples include an Autonomous Close Assistant that shortens financial close by automating journal entries, reconciliations and error resolution, and procurement-focused assistants such as JAI that embed AI directly into sourcing and spend workflows. SAP is also experimenting with Physical AI. In an active logistics warehouse, SAP and robotics software company Cyberwave have deployed fully autonomous robots that handle box folding, packaging and shipping. These robots connect to SAP’s cloud-native logistics execution platform and an Embodied AI Service that translates warehouse tasks into robot commands, illustrating how the same autonomous enterprise platform can drive both digital and physical execution.
Joule Studio AI: An Agent Factory for the Autonomous Enterprise
To operationalise this vision, SAP is turning Joule Studio into an AI agent factory for enterprises and partners. Joule Studio 2.0 lets teams identify, design and build agents for specific business outcomes, spanning no-code and pro-code development. In demonstrations, a consulting agent can detect pricing and purchasing issues, estimate margin impacts and then propose new agents as solutions, generating requirements, technical specifications and workflow logic. The latest managed version of Joule Studio removes infrastructure friction by running entirely on SAP-managed environments with built-in audit logging and enterprise controls. It supports tools like Cursor and frameworks such as AutoGen and LlamaIndex, and offers access to SAP’s own foundation models, including SAP Domain Models tailored to S/4HANA and Ariba. Integration with Claude Code and other frameworks allows Joule Studio AI users to build governed AI agents ERP teams can trust, while the Agent2Agent protocol lets third-party agents invoke Joule agents within live processes.
Ecosystem, Acquisitions and the Road to the Autonomous Enterprise
SAP’s push toward an autonomous enterprise platform is reinforced by acquisitions and partnerships aimed at filling critical gaps. Master data management specialist Reltio strengthens SAP’s ability to provide a single, trusted business context for AI agents, while Prior Labs brings frontier AI research capabilities into the fold. Partnerships with firms such as Palantir and Accenture help customers connect non-SAP data and processes, ensuring that AI agents can act across heterogeneous environments. On top of that, SAP is introducing new interfaces such as Joule Work to let users interact with enterprise automation through conversational and task-focused experiences instead of navigating multiple applications. Industry-specific autonomous products, like those developed with an energy company for offshore operations, embed sector logic and regulatory constraints directly into AI workflows. Together, these moves signal that SAP sees its future not just in software, but in orchestrating governed AI agents that run mission-critical business operations end to end.