SAP CTO discusses enterprise AI agents and tabular models
In an interview with The Rundown AI conducted in partnership with SAP, CTO Philipp Herzig said AI coding tools simplify application creation but do not remove the complexity of operating business-critical software. He discussed SAP’s enterprise-agent approach, including Joule Studio, Knowledge Graph context, Company Memory, permissions and human approvals, as well as controls for AI workload costs. Herzig also discussed SAP-RPT-1.6 for tasks such as payment-delay prediction and cash-flow forecasting. He said the model offers “Explain by Feature” to show influential variables and “Explain by Example” to show similar historical records used as evidence.
Herzig’s account highlights the business context, trusted data, governance and security requirements he says are necessary for enterprise AI. His claims about accuracy, costs and financial benefits were presented in a partnership interview and were not independently verified in the supplied report.
AI
SAP's Philipp Herzig on AI built for business data
What the "SaaS is dead" debate misses, according to SAP's CTO
Melina Panitsidis
October 25, 2026
#SAP#Joule Work#Joule Studio#SAP Knowledge Graph#SAP Company Memory#SAP-RPT-1.6#Prior Labs#Google#NVIDIA