SAP advances its autonomous enterprise push with smarter, governed Joule agents
SiliconANGLE Kyt Dotson
SAP added new AI tools for its Joule assistant and business data platform. The twist: it wants agents to work across systems, but only through governed doors.
Based on reporting by SiliconANGLE, Kyt Dotson — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
SAP used its TechEd conference to push harder on the idea of an autonomous enterprise, announcing new AI features across Joule and its Business AI Platform. The company’s pitch is not just more chatty assistants. It wants agents that can gather context, make decisions and carry out multi-step work inside enterprise systems, while still being tightly governed.
That governance piece is doing a lot of work here. SAP said its new orchestration layer will let assistants and agents invoke Joule Agents through a controlled gateway. Ted Way, SAP’s vice president and chief product officer for business AI product engineering, put it bluntly: open protocols still need “governed doors.” The company also wants customers’ own agents to cooperate with SAP’s business agents, which makes sense until you remember that the minute software starts acting like an employee, someone has to decide what it’s allowed to touch.
SAP also added support for n8n workflows through Joule Studio. That matters because n8n is deterministic automation, not a probabilistic model. An agent can trigger it, but the workflow itself follows exact rules and exact steps. For compliance-heavy work, that’s the point. SAP is trying to blend flexible AI triggers with rigid procedures, so a model can decide when to hand off and a prescribed workflow can handle the part that absolutely must happen the same way every time.
On the data side, SAP expanded Business Data Cloud Ontology Manager in SAP Data Cloud so agents can better understand how applications and facts relate to one another. It also said the recent Reltio acquisition adds multidomain data management for customer, supplier and product records across SAP and third-party systems, while the Dremio acquisition expands lakehouse capabilities with native support for Apache Iceberg and Apache Polaris and a federated query engine.
Then there’s the model work. SAP introduced SAP-RPT-2.0 for tabular data, with anomaly and outlier detection, and SAP-RPT-TS for forecasting time-indexed metrics. Way’s line was simple: not every enterprise problem is a language problem. Fair enough. A lot of business still lives in tables, not paragraphs, and SAP is betting that the boring data gets the smarter tooling.
My take — AI-written commentary, not fact-checked reporting
SAP is making the right move by treating autonomy as a permission problem, not a demo problem. The industry keeps selling agents as if they should roam free, which is a charming idea right up until they meet payroll, suppliers or a regulator. Open protocols are useful; open season is not.
Read more about this at: SiliconANGLE