SAP has thrived through multiple technology cycles by focusing on solving enduring business needs like finance and supply chain management. While the underlying technology evolves from mainframes to AI, the customer's need for business outcomes remains constant, making this focus the key to longevity.
While AI proofs-of-concept are easy, SAP's CTO states the real engineering hurdle is scaling reliably. The complexity lies in managing thousands of APIs, handling massive document volumes, and applying granular, user-specific context (like regional policies) consistently and accurately.
Enterprise AI agents create a compounding improvement loop by capturing unwritten "tribal knowledge" when they request human input. This process, termed "agent mining," records these decision traces and context, feeding a data flywheel that continuously refines the agent's autonomous capabilities.
SAP’s CTO views AI not as a feature but a fundamental architectural shift akin to the cloud transition. It requires re-engineering software at three levels: creating dynamic 'Generative UIs', automating 'Business Processes' with agents, and building a unified 'Data Layer' to power intelligence.
AI will automate mundane data collection in functions like finance and HR. This won't eliminate jobs but rather up-level them. Employees will transition from performing repetitive tasks to supervising AI agents, focusing on higher-value strategic thinking, scenario analysis, and decision-making.
Building reliable AI agents requires a developer mindset shift. The most critical task is not writing the agent's code but creating robust evaluations ('evals') that define and verify the desired business outcome. This makes a test-driven development approach non-negotiable for enterprise AI.
LLMs fail at core enterprise tasks like demand forecasting on structured data. SAP is developing "Relational Pretrained Transformers" (RPTs) to apply the foundation model concept to tabular data. This aims to democratize predictive modeling, which currently requires specialized data scientists and doesn't scale.
While AI pushes software toward consumption-based pricing, SAP employs a hybrid model. The CTO explains that enterprise customers are not ready for pure consumption as they require budget predictability and are not yet fully trusting of AI outcomes, forcing a gradual transition away from seat-based licenses.