
Meet the Authors
The strength of the ERP foundation now sets how far customers can take AI assistants and agents, according to SAP’s Maura Hameroff.
ERP modernization prepares customers for AI agents when it reworks their data and processes, and Hameroff outlines what that readiness looks like.
Planning AI, business transformation, and ERP on one roadmap helps customers prepare to manage AI agents operating across the business at scale.
“The biggest dividing line we’re seeing is whether modernization changed the underlying data and process landscape, or simply moved it,” said Maura Hameroff, chief marketing officer for SAP Cloud ERP Private and RISE with SAP.
Customers that have modernized their data and processes can give AI agents the business context to act reliably. Those that only moved their systems find agents cannot deliver the same business impact.
In an interview with SAPinsider, Hameroff described how customer priorities have shifted since SAP Sapphire in May and what that means for building an ERP foundation for AI.
AI Use Cases Have Moved From Planning to Production
The fundamental priority has not changed since May, Hameroff said: customers are still focused on business outcomes, “not experimenting with AI simply to check a box.” What has changed is that they are moving from planning into execution, often in finance and procurement, and increasingly see that AI and ERP modernization have to advance together.
Once customers look seriously at assistants and agents, the foundation beneath them becomes much clearer, she explained. “Data has to be consistent, processes integrated, and business context available if AI is going to do more than generate an insight and actually help drive an action.” The foundation is adding urgency to ERP modernization and a clean core, she added, and “the further customers progress in that journey, the more value they can get from AI.”
What Separates Customers Who Are Ready for Agents
Readiness for agents, in Hameroff’s view, depends on what modernization actually changed. Customers that invested in a modern core, harmonized their data, and connected their business processes are “in a much stronger position to put assistants and agents to work,” she said. Customers struggle, she cautioned, “when modernization has effectively been a lift-and-shift exercise without addressing the underlying data and process landscape.”
At enterprise scale, an agent has to understand the business process, the rules and controls that apply, what actions it is permitted to take, and when human review is required. With that understanding in place, customers can “move from AI that recommends to AI that can responsibly help execute,” she explained.
Accenture, which joined Hameroff at SAP Connect to discuss building an ERP foundation for AI, is her example of a customer that tied modernization to business change. Its experience shows that “technology modernization and business transformation need to happen side by side,” she noted; linking business objectives to the technology plan helped it prioritize and keep the business engaged.
Over successive upgrades, Accenture “built the discipline to keep its ERP environment current, which is now giving the business a stronger foundation for AI,” she observed. It is bringing Joule into master data analysis and preparing a Joule Studio agent for treasury.
Her advice follows from that example: “don’t separate the AI strategy from the ERP strategy.” Customers should start with the business value they want to create and modernize the foundation at the same time, she said. “You don’t have to transform everything at once,” she added, “but the AI, business transformation, and technology roadmaps need to move together.”
What SAP Is Building and What Customers Should Prepare For
Hameroff described SAP as continuing to “bring intelligence more deeply into core business processes” while strengthening governance, compliance, and industry capabilities. The common thread, she said, is that “AI is becoming part of the business process rather than sitting alongside it,” and SAP Cloud ERP Private 2025 FPS02 continues building out the foundation that more autonomous operations require.
In order-to-cash, agents can help teams identify fulfillment issues, receive suggested resolutions, and update sales orders within their workflow. In quality management, Joule and SAP Document AI can help extract and match certificate data and create pre-filled defects for review.
Feedback from early adopters is helping SAP focus on what it takes to “provision, manage, govern, audit, and connect agents when they are operating at scale across a business,” she noted. The shift customers should prepare for, she advised, “is from managing individual AI use cases to managing agents operating across the business.”
As agents take on more activity, third-party connectivity, security, and human oversight grow in importance, Hameroff observed, and “that makes the foundational work customers are doing now even more critical.”
“The autonomous enterprise isn’t going to arrive through one AI breakthrough or one software release,” she said. It will come from bringing trusted business data, integrated processes, governance, and intelligent capabilities together, and the organizations doing that work now “will be much better positioned to take advantage of what comes next.”
What This Means for SAPinsiders
Plan the AI and ERP roadmaps together. Hameroff’s advice ties AI plans to the business value they should create and the foundation they depend on. Customers gain clearer priorities and keep the business engaged.
Test whether modernization changed the data. A migration that left data fragmented and processes unconnected gives agents little context to act on. Customers that check early see where agents will lack context before they deploy.
Prepare to manage agents across the business. Provisioning, auditing, third-party connections, and human oversight become standing work as agents take on more activity within processes. Customers that put those controls in place now build the foundation Hameroff calls critical to what comes next.




