![](\"https:\/\/cdn.pblzr.de\/dacbb27c-1270-4041-b681-e2b95f06f8a1\/2026\/10\/sap-logo-400x300_1_TEXT.jpg\"){kind=logo}
<\/p>\n<\/p>\n
“It\u2019s time to show you the autonomous enterprise in action. After all, there are already impressive customer examples of this,” said Christian Klein, Chief Executive Officer of SAP, in his opening remarks at the start of the event.
Christian Klein reminded the audience that SAP had presented the autonomous enterprise at the SAP Sapphire customer conference in May 2026 as \u201cthe beginning of a new era.\u201d Since then, SAP has supported hundreds of customers in implementing Joule Work as an AI user interface for the autonomous enterprise.<\/p>\n
\u201cWe\u2019re making Joule Work available to all customers later this month,\u201d announced Christian Klein, who described the desktop app as a new user interface for AI deployment. For decades, users had to learn how SAP solutions work\u2014with Joule Work, SAP is now learning how people work. \u201cStarting now, the more than 400 million SAP users can communicate with SAP in their own language. Joule Work then takes care of the work.\u201d<\/p>\n
Christian Klein described the three cornerstones on which Joule Work is based: Dialogs, in which users ask questions and delegate tasks to agents; Spaces, where content, insights, and shared documents are displayed; and the Joule Studio solution, which allows customers to extend Joule or develop customized agents for their own processes. <\/p>\n
SAP\u2019s comprehensive business and data context is the \u201csecret sauce\u201d here. This context includes the SAP Knowledge Graph, which maps more than seven million data fields, 50 million tables, over 50,000 public and private APIs, 400 data products, and extensive process knowledge. \u201cThat\u2019s why Joule doesn\u2019t just offer advice,\u201d he said.<\/p>\n
SAP\u2019s 110,000 employees are already using Joule for hundreds of thousands of tasks every day. \u201cWe\u2019re already seeing productivity gains of over 20% in key areas such as human resources, finance, and procurement. For more complex analysis and planning tasks, the growth rates are even higher.\u201d<\/p>\n
Christian Klein then discussed the SAP Autonomous Suite, which forms the second tier of the autonomous enterprise. The suite initially covers five areas: finance, procurement, supply chain, human resources, and customer experience. He explained that the SAP Autonomous Suite is designed for tasks where a good answer is not enough\u2014it\u2019s always about \u201cthe right answer.\u201d<\/p>\n
\u201cLLMs are based on probabilities and therefore only provide the answer that is most likely,\u201d said Christian Klein. He contrasted this probabilistic approach with SAP\u2019s integrated agents. Since the agents operate within deterministic application workflows, they use the same data, follow the same rules, and generate the same audit trail as the applications themselves. Customers can control the degree of autonomy and adapt it to the respective process. For high-risk tasks, decisions must be approved by humans, while routine tasks can be automated to a greater extent.<\/p>\n
According to Christian Klein, SAP has developed over 20 Joule Assistants and more than 200 Joule Agents since SAP Sapphire. Another 400 agents are scheduled to be deployed by the end of the year. \u201cWhat really matters, however, is not the number of agents, but the business results achieved with them,\u201d he emphasized.<\/p>\n
\u201cThe reason these agents understand your business so well is the SAP Business AI Platform,\u201d explained Christian Klein. \u201cThe platform provides customers with the best Frontier AI model for every agent-based use case, gives agents direct access to the most comprehensive business context, and offers SAP-managed governance.\u201d<\/p>\n
Manoj Swaminathan, President and Chief Product Officer for the SAP Autonomous Suite, followed up on Christian Klein\u2019s remarks and described the SAP Business AI Platform as the infrastructure that lays the groundwork for the autonomous enterprise. The platform offers three key capabilities: developing agents, using business data as context for these agents, and controlling how agents function.<\/p>\n
Manoj Swaminathan explained how the context layer ensures that agents achieve more accurate results. Joule Studio enables no-code development and professional development with access to over 70 models. This strategy reduces vendor lock-in and token costs.<\/p>\n
\u201cJoule Studio is directly integrated into our context layer,\u201d he explained. \u201cAs a result, the agents developed are more precise, can be implemented more quickly, and can be executed more efficiently. Our knowledge graph is the map that includes all processes, data flows, and business tools. The agents find the right process among many thousands and select the right information from over seven million data sources. The SAP Business Data Cloud applies this context to all of your data\u2014a unified view that incorporates processes, policies, and logic\u2014to make it meaningful.\u201d<\/p>\n
Manoj Swaminathan emphasized that AI cannot eliminate the problems caused by a flawed data model. To enable agents to understand customers\u2019 business rules, the SAP Domain Models are trained using SAP code. For forecasting, SAP provides two engines based on customers\u2019 business processes: the SAP-RPT model family, which has been trained using SAP business data, and the tabular AI model from Prior Labs.<\/p>\n
Jan Gilg, Global President of SAP Customer Success & Americas, presented the fictional startup \u201cJerseys from Jersey\u201d during his portion of the opening keynote and invited a number of customers on stage to answer his questions and evaluate the strategy for the autonomous enterprise from their perspective.<\/p>\n
Among these customers were the pharmaceutical company Novartis, which is using Joule Sourcing Agents in a pilot project for bid analysis and negotiation support, and the food manufacturer Morgan Foods, which reported on the transformation of its supply chain with SAP S\/4HANA and SAP Integrated Business Planning. Also represented were the Nestl\u00e9 Group, which highlighted the importance of a unified digital core, and SAP customer and partner PwC, which demonstrated the business benefits of the SAP Industry AI portfolio.<\/p>\n
Jan Gilg summarized these customer testimonials and reiterated that the individual elements of the SAP Autonomous Suite are more than \u201ca collection of functional components. It is a system that holds everything together and provides connectivity, context, and control.\u201d<\/p>\n
Jan Gilg also had several product announcements in store, including the launch of the integrated payment service SAP Pay, which can automatically execute and reconcile supplier payments as soon as an invoice is approved. He also announced the planned acquisition of TechWolf, a Belgian startup that has developed a workforce intelligence solution to close skills gaps.<\/p>\n
Marielle Ehrmann, Chief Security Officer and member of the SAP Supervisory Board, then addressed the topics of trust and governance. She reported that ransomware attacks, which used to take attackers eight hours to carry out, can now be executed in just 22 seconds with the help of AI. She also pointed out that while 83% of companies use AI on a daily basis, only 13% have real insight into the AI they deploy.<\/p>\n
\u201cThe greatest threat to companies is not malicious AI, but AI that is not managed,\u201d explained Marielle Ehrmann. To defend against this threat, each agent should be assigned a verified administrator, a risk level, and clearly defined access rights. These should be supported by zero-trust checks of identity, purpose, and data sensitivity, as well as by human oversight.<\/p>\n
She emphasized that the SAP Business AI Platform incorporates legal regulations and localization requirements, enforces user permissions, ensures that customer data is not used to train foundation models, and supports three levels of control: human-in-the-loop approval, human-on-the-loop monitoring, and the human-in-command model, in which humans are the final decision-makers. \u201cThe more autonomously an agent acts, the stricter the control measures must be,\u201d explained Marielle Ehrmann. She described this principle as \u201cfreedom within clearly defined boundaries.\u201d<\/p>\n
In closing, Christian Klein took the floor once more and emphasized that trust is the foundation of SAP\u2019s business operations\u2014not only over the past 50 years, but also in the future. “We provide the most trustworthy AI with the best governance,” he said. “However, the use of AI also requires comprehensive change management so that it can deliver the desired results for your company.”<\/p>\n
This is where SAP\u2019s solutions for business transformation come into play. SAP Signavio measures the value created by processes before and after a change. This information can be used to justify token costs. With SAP LeanIX, companies can identify and monitor AI agents, while WalkMe promotes user acceptance and trust. Christian Klein also highlighted the tools for agent-driven migration. In his view, these solutions have meant that companies no longer have to incur expenses in the millions for migration projects.<\/p>\n
\u201cThis is no longer just the beginning of a new era. This is a new era,\u201d Christian Klein told the audience. He explained that a significant shift is taking place among SAP customers: \u201cFrom manual steps to connected workflows, from reactive decisions to predictive insights, from manual spreadsheet processing to Human-in-the-loop and human-on-the-loop\u2014all of this is based on our AI platform, which ensures that AI remains precise, controlled, and scalable.\u201d And above all, according to Christian Klein, \u201call of these possibilities are already available. This is the autonomous enterprise.\u201d<\/p>\n
– – – – –<\/p>\n