At the beginning of his session at the recent Next Generation SAP Enterprise Architect Learning Forum, Johannes Euler, Head of AI for SAP Enterprise Architects, posed a serious question to the audience: Is SaaS dead? 

The audience hesitated. A few hands went up in agreement. Most stayed down. 

“I like your optimism,” Euler joked, before reframing the issue. The real question isn’t around the death of SaaS but how EAs adjust to the major AI transformation underway. It’s a deeper shift in how enterprise systems create value and how EAs are now on the front line of that change. 

Thankfully, Euler outlined five important shifts EAs need to pay attention to this year to help make the AI transition easier. 

1. From AI in Applications to AI on Applications 

For years, incremental AI features have been layered into ERP and supply chain systems, but that model is changing. 

“What we are seeing with AI is that it is also becoming a commodity,” said Euler. “That leads to the shift that we should not just build AI within the apps, but we have an agentic layer on top of the apps.” 

This new layer changes everything. Instead of enhancing individual applications, AI now orchestrates across them. For EAs, that means designing for AI as a control plane, not just a capability. 

2. The End of Application Boundaries 

Traditional enterprise architecture has long been application-centric. But agentic AI is exposing the limits of that model. 

“When you look at supply chain, it’s not just one system,” Euler said. “It’s maybe also some other vendors or custom solutions that are acting in this space.” 

The result is a clear shift from app boundaries to landscape-wide intelligence, which is why organizations must now architect for cross-system workflows, end-to-end business processes, and interoperable AI agents. The future isn’t about optimizing systems in isolation; it’s about enabling intelligence across the entire landscape. 

3. Value Can’t Wait for Modernization 

For years, organizations were told to modernize before unlocking the value of AI. But it appears that strategy might soon be outdated. 

“A lot of companies have found a workaround,” Euler said. “They have found an easier way to AI that maybe isn’t as enterprise-ready, but it somehow gets you there.” 

Even as enterprises move toward platforms like SAP S/4HANA, business leaders are demanding immediate returns. This is creating a new approach. 

“We have to offer you value while you’re modernizing and not just wait for you to be modern,” said Euler. 

For architects, this means designing parallel value streams, where AI delivers outcomes during transformation, not after it. 

4. From Human-Led to Agent-Led Transformation 

Enterprise transformations have traditionally depended on armies of consultants and systems integrators. That model is evolving fast, according to Euler. 

“Climbing this mountain is quite intense,” he said. “You have a lot of project effort and a lot of SI armies joining you.” 

Now, agentic AI is starting to take on part of that burden as agent-led migrations increase speed and quality. The implication is not a replacement but an augmentation. Architects must design for hybrid execution models, where humans guide strategy and agents accelerate delivery. 

5. From Shipping Features to Delivering Value 

One of Euler’s most striking shifts had less to do with technology and more to do with operational mindset. 

“There is a big cliff between what’s actually there and what you are actually using,” he said, describing the gap between delivered AI capabilities and real-world adoption. That gap has multiple layers: 

• Procurement friction
• Activation complexity
• Organizational readiness  

For EAs, success is no longer deployment. It’s adoption and impact. 

“We have to shift away from ‘we have shipped a capability’ to celebrating the actual value realization,” Euler said. “Users are actually using it, which makes their life more productive.” 

The New Mandate for EAs 

Taken together, these shifts redefine the role itself. Enterprise architecture is moving to the following: 

• From application-centric to AI- and data-centric
• From system design to value orchestration
• From governance to enablement at scale 

“I think we as EAs are right in the front line to translate today’s world into what’s coming next,” said Euler, who added that this responsibility comes with a new playbook: 

• Prioritize value over capability. “Look always at the value when designing use cases.”  
• Adopt before you build. “See what you can take off the shelf, then extend, and only then build from scratch.”  
• Prepare for scale early. “You must secure that your enterprise is ready, not when the CIO calls you because then it’s too late.”  
• Embrace data as the foundation. “AI needs this context in order to be relevant and reliable.”  

SaaS Isn’t Dead, It’s Becoming Something Else 

So, is SaaS dead? Not exactly. But it is evolving into something fundamentally different, according to Euler. 

“I decided to call it work as a service and intelligent work as a service,” he said.  

Under this new model: 

• Applications fade into the background
• Agents orchestrate work across systems
• Value is delivered through outcomes, not interfaces  

For EAs, the future isn’t about designing applications. It’s about designing how work happens in a world where agents are increasingly doing the work.