From Software Vendor to Business AI Company
In his Sapphire keynote, SAP CEO Christian Klein framed the company’s future with a stark question: will SAP still be a software company? The answer came in the form of the SAP Autonomous Suite, a reimagined ERP stack that places AI agents at the center of business execution. Rather than treating AI as an add-on, SAP is building what it calls a “system of execution” where applications, data, and process logic feed a layer of intelligent agents that can reason, recommend, and act across finance, spend, supply chain, HR, and customer operations. This marks SAP’s explicit repositioning as a business AI company, with software becoming the substrate rather than the product. The move is designed to shift value from static record-keeping to continuous orchestration of outcomes, suggesting that the next phase of ERP competition will be fought over who controls the autonomous control layer above traditional applications.
Data and Context: The Foundation of the SAP Autonomous Suite
SAP’s vision for the Autonomous Suite rests on a simple premise: AI agents fail without clean, consistent, and connected data. Klein stressed that “no AI agent can compensate for a bad data landscape,” positioning SAP’s business data layer and context model as the bedrock of its agent platform. The company is extending this layer across SAP and non-SAP estates to provide a single business context for agents, rather than another patchwork of integrations. Customers can already tap into hundreds of managed data products and use a new generation agent to help design additional data products faster. SAP’s data strategy goes beyond analytics; it aims to encode process logic, master data, policies, and transactions into a governed context that agents can safely act on. This reframes ERP from a system of record into a system where data interoperability and governance directly power enterprise automation.
Joule Studio 2.0: An Agent Factory for Enterprise Automation
At the center of SAP’s business AI push is Joule Studio 2.0, presented as an “agent factory” for enterprises. CTO Philipp Herzig demonstrated how the platform helps organizations identify, design, and build agents geared to specific business outcomes, starting with a consulting agent that pinpointed a pricing and purchasing issue with a substantial margin impact. Joule Studio 2.0 then generated the requirements, technical specifications, workflow logic, and orchestration needed to implement a sales pricing validation agent. Unlike generic prompting tools, Joule Studio 2.0 is intent-based and model-agnostic, but deeply grounded in SAP’s business data and process semantics. It can target both SAP and third-party environments, while embedding governance, controls, and evaluation capabilities. The result is a more industrialized approach to enterprise automation, where organizations can move from idea to governed, production-grade agents that are designed to interact with existing ERP landscapes rather than bypass them.
Packaging the Autonomous Suite Around Agents and Assistants
SAP is not just launching isolated AI features; it is reorganizing its portfolio around agents and role-based assistants under the Autonomous Suite banner. Product and engineering leadership highlighted that the company has already built hundreds of agents and dozens of assistants across four major business processes, with more planned on a rolling basis. These assistants can be triggered by humans or by systems, are mapped to specific roles, and are instrumented to track business impact. Crucially, they can be extended with rules, workflows, and code using the same experience provided in Joule Studio 2.0. For finance, that means assistants for close and controlling; for spend, sourcing and buying; for supply chain, need-to-deliver workflows; for HR, recruiting and career development; and for customer-facing functions, sales, service, offer, and marketing scenarios. The outcome is a suite where end-users increasingly work through autonomous agents rather than navigating traditional application screens.
Physical AI: Autonomous Warehouse Robots as Proof of Concept
SAP’s business AI story extends beyond digital workflows into the physical world, with autonomous warehouse robots operating in a live logistics site. Working with robotics software company Cyberwave, SAP has deployed AI-powered robots that perform box folding, packaging, and shipping tasks in an SAP-run warehouse. The deployment relies on SAP LGM as the digital backbone, together with SAP BTP and Physical AI capabilities, enabling robots to be introduced quickly, operate reliably, and scale as processes evolve. Cyberwave’s platform uses demonstrations from human operators to train Vision-Language-Action and reinforcement learning models so robots can generalize across different box types, layouts, and order mixes. This “demonstrate and deploy” model replaces brittle, hand-coded scripts with adaptive policies that evolve with floor conditions. For ERP practitioners, the project signals that SAP sees autonomous warehouse robots and Physical AI as repeatable patterns, not one-off showcases, pushing enterprise automation into production logistics environments.