Five takeaways from SAP TechEd Berlin

SAP’s TechEd conference this week offered a few surprises, but its focus, predictably, could be spelled in two letters: AI.

Artificial intelligence was everywhere, as the more than 2,800 in-person delegates and the 20,000-plus online attendees quickly discovered. However, the technology was woven into the various components in ways ranging from the expected to the novel.

Here’s a look at five things that caught our eye.

The rise of zero copy

SAP is becoming more open according to Muhammad Alam, SAP Executive Board Member in charge of product and engineering. He highlighted the company’s link-ups with Databricks, Google Cloud’s BigQuery, and the newest addition, Snowflake, that allow zero-copy bidirectional data sharing between the three platforms and SAP Business Data Cloud (BDC). And those three are just the start, he said: SAP has more integrations planned, though he declined to reveal with whom.

Certification shakeup

Passing tech certifications used to involve multiple-choice marathons that basically tested memorization skills. SAP is changing that testing model over the next year, instead requiring candidates to address real-world scenarios and allowing them to use available reference materials, including AI, to solve the problem, just as they would in the office.

The rationale is that only a prepared candidate would know which references to use in a given situation, so this methodology more realistically tests their skills as they’d be used in business. Six exams are available now, with the remainer of the portfolio to be transitioned in 2026.

AI everywhere

Every enterprise is becoming a data company, Alam said during his keynote, and every user experience is becoming AI driven. “These shifts lay the foundation for what comes next, the era of agentic AI, where AI moves from just being a tool to becoming your trusted teammate,” he said.

To power that shift, SAP plans to add even more AI agents to the 20-plus already available. They’ll be added to everything from development tools to business processes, and to help users adapt, the company has committed to equip 12 million people worldwide with AI skills by 2030.

A new foundation model

SAP launched what it described as “the first foundation model built specifically for structured business data”, a giant model pretrained to understand table-based business data across many predictive tasks. It’s dubbed sap-rpt-1, (rpt stands for Relational Pretrained Transformer), and the company said that it offers in-context learning, so It doesn’t require task-specific training or fine-tuning.

“You provide labeled table row examples of your predictive task in-context, and the model delivers a highly accurate prediction instantly,” SAP CTO Philipp Herzig explained. “One single model, serving a huge variety of enterprise prediction tasks, across finance, supply chain, HR, and more, simply by each time seeing examples from a different domain in its context.”

And now for something completely different

Two somewhat off-the-wall announcements, both of which involved AI to some extent, showed SAP spreading its wings.

First, the company is expanding its Embodied AI ecosystem, working with robotics companies to, it said, drive the future of physical AI. It is not, Herzig emphasized, getting into the hardware business, but rather is developing software to allow its customers to integrate robotics into their business functions.

The second area showcased SAP and IBM’s partnership in quantum computing. Again, SAP is not considering the hardware side, but is developing algorithms and business processes, and, said Herzig, “to kind of interweave quantum computing as an additional compute paradigm into the business process software” without having to know anything about the technology.

“We believe quantum will join classical and AI compute in your stack, and we are embedding it into the processes and apps you already use,” he said, “so it just shows up in the workflows of your enterprise.”