A Note from the CEO

A Journey Through the Evolution of Enterprise Systems

I entered the world of enterprise software through SAP projects at a time when three-tier client/server architectures were widely considered the safe and proven end state of enterprise computing.

When I began my PhD at the Hasso Plattner Institute in 2006, working with Professor Hasso Plattner, it quickly became clear that this supposed “end state” was already beginning to collapse under the weight of rapidly growing data volumes and increasing system complexity.

In December 2006, Jim Gray delivered his famous talk:
“Tape is Dead, Disk is Tape, Flash is Disk, RAM Locality is King.”

It was more than a memorable line. It was a clear signal of where enterprise systems needed to go—and it helped inspire what later became SAP HANA.

Around the same time, Workday was building its in-memory OMS. The underlying idea was remarkably similar: use memory as the central foundation of the system. The difference was that Workday built this approach without legacy constraints.

Looking back, that period marked a true inflection point: the emergence of systems capable of handling mixed workloads while extending the life and relevance of traditional systems of record.

Years before Gartner introduced the term HTAP, I described the concept in 2010 as OLXP in my paper “A Case for Online Mixed Workload Processing.”

From Theory to Enterprise Reality

In 2013, I joined SAP to help translate these architectural ideas into real enterprise applications. At the time, the project was called Simple Finance. The name S/4 Finance came later. The goal was straightforward: leverage the new capabilities of HANA to rebuild finance processes in a way that truly matched what the database could do. This work eventually became one of the foundations of S/4HANA. I later summarized these ideas in my book “SAP S/4HANA Finance: An Introduction.”

Technologically, the step was a success.

Performance improved dramatically.

Data models were simplified.

Real-time reporting became possible.

Tasks that were previously slow or impossible suddenly became feasible.

But something else became equally clear:

Cloud-native is not simply a hosting decision.

It fundamentally changes how software is built, operated, and evolved. At the same time, the market moved strongly toward point solutions in HR, CRM, and other domains. And another lesson kept repeating itself:

Transactional systems are designed to record and reconcile data. They are not designed to understand the flow of work.

Seeing Enterprise Systems from the Customer Side

Later, during my time at McKinsey, I experienced enterprise systems from the customer perspective. That experience changes how you think. Architectural debates become less theoretical. Everything is measured against business value and the speed of change. You also begin to see the real cost of complexity:

The long tail of legacy decisions

Fragmented integration landscapes

Frozen processes

The organizational consequences of systems that cannot adapt

Innovation rarely fails because people lack ideas. It fails because change becomes too expensive and too slow.

Learning the Discipline of SaaS

In 2019, I joined Workday. Seeing a true SaaS company from the inside was a completely new perspective—and it filled an important gap in my experience.

I had the privilege of learning directly from Petros Dermetzis, part of Workday’s founding team, and understanding what “Power of One” really meant in the early Workday years. It was not a marketing slogan. It was a strict product and engineering discipline:

one code line

one release for all customers

one security model

one consistent data foundation

This discipline creates enormous leverage. It makes platforms easier to operate, easier to evolve, and easier to trust. But it also taught me something else: “One” does not mean an enterprise has only one system. It means that whatever you build as a platform must be coherent, composable, and capable of connecting cleanly to the rest of the landscape. And it reinforced something I had come to believe strongly:

Enterprise software only works when it is people-centric, and when workflows are treated as first-class citizens rather than afterthoughts.

The AI Inflection Point

In spring 2023, I met Rainer Glaess, the visionary founder of GK Software. He introduced me to the Newwork project, and at that moment the direction suddenly became clear—just as generative AI was breaking into the mainstream. For me, it was another inflection point. Systems of record have been optimized for decades. SaaS has proven how to run products continuously at scale. But the interaction model and execution model of enterprise systems had barely changed.

Most enterprise software still relies on forms and menus, even though real work in organizations is:

event-driven

cross-functional

dynamic and fast

Generative AI opened a new chapter—but not in the way many people first imagined. It is not about placing a chat window in front of legacy systems. In the enterprise, AI only becomes meaningful when it operates:

inside governance

inside policy

inside an audit trail

inside real operational execution

AI needs context. It needs deterministic guardrails. And it needs the ability to trigger and orchestrate work, not merely comment on it.

Building the Execution Layer for the Autonomous Enterprise

That realization became the foundation for NEWWORK Software. Our goal is to go beyond ERP and beyond silos. We merge data, solutions, and collaboration into a unified execution layer. The objective is simple but powerful: Control the flow of work—so the system can do the work. At the center of this approach is FLOW.

FLOW is the backbone that:

Turns events into actions

Moves work seamlessly across functions

Ensures responsibilities follow the organization as it evolves

Makes automation reliable and governed

This is where everything I have learned over the last two decades comes together:

real-time systems

cloud platforms

measurable business value

people-centric workflows

and now AI as a true part of the operating model