SAP Details Autonomous Enterprise Strategy at TechEd Conference

On November 22, 2026, SAP detailed its “Autonomous Enterprise” strategy at the TechEd developer conference in Berlin. The framework integrates AI agents within the SAP system to perform analyses and execute tasks, with the company currently refining methods for cost control, oversight, and model management within its SAP Business AI Platform.

Managing AI expenses remains a primary challenge for companies deploying autonomous agents. According to Jonathan von Rüden, chief AI officer at SAP, the company assists clients in estimating consumption based on employee counts, process frequency, and data volume, as Computerwoche reported.

“Ultimately, a large part depends on usage. However, budgets and limits can be set to control costs,” von Rüden told Computerwoche. Within Joule Work, administrators can implement daily usage limits for individual employees, with progress tracked via percentage indicators. If a budget is exhausted, staff may request additional allocations for critical tasks.

To optimize spending, von Rüden suggests that companies set default models that balance performance with price, rather than defaulting to the most expensive options. SAP is also testing automatic model routing internally, which would eventually assign the most efficient model to a specific task without user intervention.

Managing Agent Proliferation

Not every user receives automatic permission to build agents in Joule Studio. IT departments assign roles to control access, and SAP itself limits deployment to employees who have completed specific training.

“First of all, not every user automatically has permission to use Joule Studio. This role is assigned by IT,” von Rüden told Computerwoche. Individual customizations often start as simple skills, which can be shared with colleagues for testing before IT reviews them for broader use.

The SAP AI Agent Hub provides a centralized inventory to track the lifecycle of these tools. It allows organizations to monitor which agents are under review, approved, or rejected, and facilitates the planning of decommissioning dates for outdated systems.

Operational Oversight and Future Scaling

For reliability, SAP integrates development, monitoring, and process analysis. By connecting to SAP Cloud Application Lifecycle Management, companies can examine the specific decisions an agent makes and compare its performance against existing process paths.

While businesses currently rely on a mix of manual governance, SAP is introducing tools to standardize policy enforcement. The SAP Agent Gateway provides free, controlled communication between Joule agents and external tools.

For more complex requirements, such as restricting actions based on transaction amounts, the premium SAP Enterprise AI Gateway allows for broad policy enforcement.

Regarding the future density of these systems, von Rüden offered a benchmark for implementation.

“A rough rule of thumb could be: Let’s take a company that uses roughly a hundred different SaaS applications, whose tasks are initially covered by 10 to 30 agents,” he told cio.com.

As technology advances, SAP is investigating models that can learn directly from feedback to adjust their own weights. “In my estimation, however, research in this area is still in its early stages,” von Rüden told Computerwoche.

Photo of author

Technology Desk

Technology Desk at Archyde.

Sports Platforms Prepare AI Launch to Combat Traffic Drop, SportsPro Says