Welcome to this exclusive interview with Philipp Herzig, chief AI officer and chief technology officer, SAP. Herzig sits down with Bob Evans to talk about the evolving role of AI in business and SAP’s commitment to shaping the future of enterprise tech.
Please note: At the time of filming, Herzig held the title of Chief AI Officer, which is why he is introduced as such. Since then, he has also taken on the role of CTO.
Highlights
Interplay Between Agents and Apps in AI (01:01)
Current AI experiences are narrow and explicit, says Herzig, and have less agency but will become more autonomous over time with advancements like reasoning models and deep learning. New enterprise apps will exist for a long time, “but the architecture and how they are designed . . . with each new step will fundamentally change.” Agents play a significant role in various business processes like claims management and financial closing and need to integrate with applications for data and functionality.
The Role of Knowledge Graphs and Neuro-Symbolic AI (03:11)
Herzig explains the role of Knowledge Graphs in guiding agents by providing business process lineage and descriptions. He compares the current state of AI agents to a team of experts working on a task, emphasizing the need for structured guidance. When large language models are combined with deterministic approaches like symbolic AI systems, better results are achieved. He draws a parallel between the evolution of AI agents and the transition from on-prem software to cloud-based SaaS applications.
Customer Concerns and the Future of SAP Applications (07:02)
AI agent innovation is a natural continuation of SAP’s applications. Herzig says that building agents outside of SAP’s applications can lead to a loss of context and integration, which is crucial for business processes. Ultimately, SAP’s focus is on end-to-end business processes: “There’s a lot of standard business processes like accounts payable or order management and a bunch of other things where you can get higher degrees of efficiency and productivity through ever more agentic workflows or agentic experiences in a standardized way . . . that resonates very well with our customers.”
AI Agent & Copilot Summit is an AI-first event to define opportunities, impact, and outcomes with Microsoft Copilot and agents. Building on its 2025 success, the 2026 event takes place March 17-19 in San Diego. Get more details.
Industry-Specific Applications and Partners Ecosystem (10:14)
SAP focuses on horizontal AI platforms for enterprise functions like finance and HR, but also builds verticalized solutions for specific industries. A strong partners ecosystem helps develop industry-specific applications that run on SAP’s platform. Herzig mentions that while SAP will focus on standard business processes, there will still be room for custom agents in specific industries.
Pricing Structure and Commercial Models (12:00)
SAP has introduced a commercial model with AI units that customers can purchase and use across various SAP products. Herzig mentions that the value of AI units is divided between SAP and the customer, with 80% of the value going to the customer. He discusses the transition to a new pricing model as AI agents become more integrated and replace traditional seat-based licenses. “What will happen? . . . Step by step, certain things, where in the past, you required a traditional seat-based license will either stay flat or will go down. Hence you need a new mechanism to. . . monetize the value that is being created through these agents and this kind of business model transformation that is going to happen.”
Evolution of SAP’s Business Model (14:58)
SAP has evolved from a classic application company to one that integrates AI, data, and platforms. SAP’s goal is to make technology disappear and focus on customer outcomes, ensuring a return on investment. Herzig concludes that higher adoption of AI in the enterprise will be achieved by making these technologies easy to consume and integrate.
Ask Cloud Wars AI Agent about this analysis
