SAP’s bold AI promises from Sapphire 2025 are finally reaching customers, but the rollout has been slower and more iterative than originally anticipated. The enterprise software giant announced Knowledge Graph, Joule Studio, and AI Agent Hub at last year’s conference, with a target of general availability by the end of 2025. While all three tools are now technically available, adoption has lagged, and SAP is already unveiling version 2.0 of Joule Studio to address early shortcomings.

What actually shipped

Joule, SAP’s generative AI assistant, and the AI Agent Hub are now generally available. However, the AI Agent Hub is undergoing a “massive revamp” with version 2.0, according to SAP executives. The Knowledge Graph, initially designed to help build Joule skills, is live and has expanded its role—it now provides contextual data to AI agents, enabling them to dynamically determine how to call functions.

Joule Studio, the low-code tool for building custom AI agents, remains in early customer adoption. General availability is now expected in the third quarter of 2026, roughly a year behind its original target. SAP’s chief AI officer, Jonathan von Rüden, acknowledged the delay, explaining that the initial version prioritized ease of use over power. “People wanted to see more pro-code flexibility,” von Rüden said in an interview at Sapphire 2026. “We had gone with a low-code approach. You could give it extension points and tools, but you couldn’t touch the core of it. Now you can build a custom agent, connect it to your own GitHub.”

Customers came with ambitious plans but needed hard rules and approval gates that the original Joule Studio did not support natively. “What people want is agentic flows with clear gates and workflows and subagents,” von Rüden added. “Old Joule didn’t provide that. Now it’s all baked together.”

What’s different in 2.0

The revamped Joule Studio addresses the gaps that held back adoption. Beyond pro-code flexibility, developers can now build with popular agent frameworks such as LangGraph and AutoGen. Agents will have native understanding of SAP’s proprietary ABAP code, business object models, and data schemas—something generic tools cannot replicate. This allows agents to work seamlessly with SAP S/4HANA, SuccessFactors, and other enterprise applications.

“The first runs were geared toward automation,” von Rüden said. “Now agents need to bring optimization and intelligence as well.” The new version introduces autonomous sub-agents that can be chained together with conditional logic, approval workflows, and escalation paths. Customers including Ericsson, Mercado Libre, and Siemens are already using Joule agents in production, handling tasks like purchase order approvals, customer service escalation, and supply chain anomaly detection.

Meanwhile, SAP is rethinking how it gets AI tools into users’ hands. Joule Desktop, launching this week, lets individual users build automations without going through IT—a bet that grassroots adoption will move faster than centralized rollouts. This mirrors a broader trend in enterprise software where citizen developers and no-code tools are empowering business users to automate routine tasks.

Background on SAP’s AI journey

SAP has been investing heavily in AI for years, but the current wave began with the launch of Joule in 2023. Initially a simple conversational interface for SAP systems, Joule evolved into a platform for building and deploying AI agents. The Knowledge Graph, announced as a foundational piece, maps relationships between business entities—customers, orders, products, employees—to give AI a structured understanding of the enterprise landscape.

The slow adoption of Joule Studio reflects a broader challenge in enterprise AI: balancing power and ease of use. SAP’s low-code approach was intended to democratize agent creation, but enterprises quickly found they needed more control over logic, data access, and security. The shift toward pro-code in version 2.0 signals SAP’s willingness to adapt to customer demands, even if it means delaying general availability.

Competitor Dynamics, Oracle, and Salesforce have also announced AI agent platforms, but SAP differentiates by integrating deeply with its own ecosystem. For customers already running SAP, the promise of pre-built connections and native data models is a strong selling point. However, the delayed availability of Joule Studio may cause some enterprises to experiment with third-party tools instead.

Key facts from the original article

• SAP announced Knowledge Graph, Joule Studio, and AI Agent Hub at Sapphire 2025, with a target of shipping by end of 2025.
• All three tools are now technically available, but adoption has lagged. SAP is releasing version 2.0 of Joule Studio to address early shortcomings.
• Joule Studio adoption was minimal because it was limited to content-based experiences; customers wanted pro-code flexibility, hard rules, and approval gates.
• SAP’s chief AI officer Jonathan von Rüden said the original architecture favored ease of use over power. Version 2.0 adds pro-code capabilities, support for LangGraph and AutoGen, and native understanding of SAP’s proprietary code and data models.
• Joule and AI Agent Hub are generally available; the AI Agent Hub is getting a “massive revamp” with version 2.0. The Knowledge Graph is live and now feeds context directly to AI agents.
• Joule Studio is still in early customer adoption; general availability is expected in Q3 2026, a year behind schedule.
• Customers including Ericsson, Mercado Libre, and Siemens are using Joule agents in production.
• Joule Desktop, launching this week, allows individual users to build automations without IT approval, aiming to boost grassroots adoption.

Implications for enterprise AI strategy

SAP’s experience highlights the tension between speed and depth in enterprise AI deployments. While vendors rush to market with generative AI features, many customers find they need customisation, governance, and integration before they can achieve ROI. The delay in Joule Studio’s full availability suggests that SAP is listening—but also that the complexity of enterprise data and workflows cannot be glossed over by low-code wizards alone.

For CIOs and IT leaders evaluating SAP’s AI roadmap, the key takeaway is that the tools are maturing but not yet mature. The promise of agentic AI—autonomous digital workers that can execute multi-step business processes—is compelling, but the path to production involves learning, experimentation, and careful governance. SAP’s decision to invest in pro-code flexibility and extendable frameworks aligns with what many large enterprises are demanding: AI that fits their existing development practices, security policies, and audit requirements.

As SAP rolls out version 2.0 of Joule Studio and continues to enhance the AI Agent Hub, the next few quarters will be critical. If adoption accelerates, SAP could solidify its position as a leader in enterprise AI. If not, the company risks losing ground to more agile competitors—or to the growing ecosystem of specialised AI agent platforms that can integrate with any back end.

Source: InfoWorld News