Wolfgang Wendt has been managing IBM’s business in Germany, Austria and Switzerland for almost two years. In total, he has been active at IBM for more than three decades in a wide variety of roles, from customer-oriented roles as Client Director to responsibility for mainframes and other hardware segments. In an interview with COMPUTERWOCHE, he looks back — and forward.
IBM is over a hundred years old; you yourself have been on board for around 30 years. How would you briefly describe the development of the company in recent years?
Wendt : We are now 114 years old, in Germany almost 115. I like to describe IBM as an ‘older start-up’ because IBM has undergone many transformations over the decades. In the beginning, everything was strongly characterized by the infrastructure, keyword mainframe. Later came the PC boom, a distributed world that we played a decisive role in shaping.
This was followed by a phase in which we focused more strongly on the service sector—data center operations, outsourcing and managed services. With the start of the cloud era, we focused more on technology again, particularly on software, but also on hardware. And around five years ago, we focused strategically on two key areas: hybrid cloud and artificial intelligence.
IBM has been working with AI for much longer than that.
Wendt: Yes, we have been focusing on AI for over 30 years. IBM was there early on, think of the chess computer “Deep Blue” or later Watson, known for its appearance on Jeopardy! With the emergence of ChatGPT a few years ago, the topic has become enormously popular. Today, it is all about using AI sensibly in both large digitalization projects and industry-specific applications.
Do you benefit from the fact that IBM has been involved with AI for a long time?
Wendt: I think so. We have a big head start here, not least thanks to the openness of our models. We have our own Granite models, but customers can also use models from the open source environment or from providers such as Meta.
The models run on-premises, i.e. independently of hyperscalers and therefore detached from the internet. This allows companies to learn their own data without any data outflow. In my view, this is one of the key differentiators.
We also follow the principle of ‘trained is better’ — in line with what DeepSeek demonstrated six months ago. We expressly support the idea of developing customer-specific models, i.e. training large language models, ‘boiling them down’ and enriching them with the really relevant data.
The inclusion of our own company data also plays a major role here in order to develop it further and make it usable — this is one of our great strengths.
Another advantage lies in our “watsonx.governance” module. It ensures the traceability of AI evaluations: Where does the data come from? How are decisions made? Who has access to it? And how reliable is it? This is another area where we clearly stand out from many of our competitors.
Our customers also use the watsonx.ai module to integrate one or more AI models from other manufacturers. We also work with well-known partners, such as SAP, which relies on our watsonx technology in its Joule assistant.
We also offer another powerful solution: intelligent orchestration via agents. We can generate these agents automatically within five minutes. But even more important is the intelligent orchestration of these agents: it enables our clients to integrate other providers such as Google, AWS or standard applications such as Salesforce via these agents and thus make targeted use of AI.
IBM used to be almost synonymous with reliability. The saying “You won’t get fired with IBM” was proverbial. Does that still apply?
Wendt: Reliability and quality are still core values at IBM. We have been supporting many of our major customers for decades. That is firmly anchored in our DNA. At the same time, we have constantly evolved.
What is important to me is that we want to provide our customers with proven technologies. The times are fast-moving, new products are constantly emerging that first need to be tested.
We take a very balanced approach here. We call it ‘Client Zero’. We test and use the technologies that we sell, such as automation and artificial intelligence, in our own company first. The focus is clearly on process and cost optimization.
When we launched Client Zero in 2022, we had targeted savings of two billion dollars. After two years, we can say that we have even exceeded the target with savings of over three billion dollars.
We want to remain a reliable technology partner—also in consulting—and, as already mentioned, together with our ecosystem partners. This includes hyperscalers as well as established system integrators.
Here too, we believe that less is more. We used to outsource on a large scale and manage complete data centers. This is no longer our core business; we spun off this area around five years ago with the spin-off of Kyndryl. Today, we are consistently concentrating on technology, software and the further development of our hardware strategy, particularly in the mainframe and storage sector.
What role does the German location play for IBM? How much development is carried out here?
Wendt: We can look back on over 60 years of development tradition in Germany. I would particularly like to highlight our IBM laboratory in Böblingen near Stuttgart. It was established in the 1950s and is still a central location with over 1,200 developers today, in a laboratory network with our facilities in the USA and other locations worldwide.
We have been carrying out both hardware and software development there for decades. For example, parts of the IBM System Z mainframe system—a key technology—were and are still being developed in Böblingen.
The location also plays an important role in the field of artificial intelligence: functions relating to data gathering, i.e. the development of a data lakehouse, are being created in Böblingen for our watsonx platform.
In addition, we opened a Quantum data center in Ehingen last October.
We are very proud to be able to build on this decades-long tradition and that technology and innovation continue to be created in Germany. Especially in the context of discussions about digital sovereignty, this is very well received by authorities. The state government is also proud that we are represented here in the “Ländle”—and this is ultimately also of great importance throughout Germany.
IBM
With AI, the mainframe is supposedly experiencing a revival. To what extent do you notice this in your business?
Wendt: Fortunately, the mainframe is experiencing a revival with almost every generation. We have now reached the 17th generation with the z17.
The great advantage of the mainframe is that sensitive data remains in a protected system. Many of our customers—especially banks, but also other transaction-oriented companies—run business-critical applications on these systems: Core banking systems, transaction platforms, credit card processing and the like.
Here in particular, it is crucial that we integrate artificial intelligence directly into these systems instead of running it as an upstream module on a separate computer or GPU. This would undermine the integral character of the mainframe.
This is exactly where our innovations come in: We have integrated AI — specifically the so-called runtime module — directly into the Telum processor. A simple example: When you pay with your credit card at the checkout, you expect the transaction to be secure and completed within seconds. AI supports this in real time — for example with fraud detection, cybersecurity checks or credit card authentication. These checks must be carried out in milliseconds. If AI were to cause a transaction to suddenly take a minute instead of three seconds, it would be counterproductive. That’s why the evaluation must take place close to the processor, i.e. directly on the chip — in real time.
I always say: if AI doesn’t find its way into our mainframe systems, then we’re doing something wrong.
AI — and generative AI in particular — also promises to take over activities from employees, which could lead to job losses. This is being strongly promoted in the USA, including by the IBM CEO. What is your position on this as Head of Germany and Head of Technology?
Wendt: We continue to recruit in a very deliberate, considered and focused manner and at the same time make changes where it makes sense. This is a normal process, partly driven by AI, partly independent of it.
At the same time, we are also creating many new, AI-oriented roles. Of course, there are also efficiency gains. In our Client Zero project, for example, the aim was to optimize purchasing and HR processes. This can mean redistributing tasks or adapting jobs.
As Managing Director of IBM Germany, I will continue to carry out selective restructuring in the future. But our primary goal is to provide our employees with comprehensive training in AI so that they are at the forefront of technology and consulting.
We don’t see AI as a replacement for people, but as support. Our aim is not to cut jobs, but to make employees more productive, more precise and more effective, with better output and a lower error rate. This is the approach we are pursuing at IBM in Germany.
Various industrial sectors in Germany are currently weakening, especially the automotive industry. How is this affecting IBM’s business?
Wendt: As IBM in Germany, we naturally feel committed to the German economy and are part of it. What is currently happening in the automotive industry is a profound transformation that we are actively accompanying and supporting.
Of course, we can see that some consulting projects are currently declining somewhat simply because many car manufacturers are implementing cost-cutting programs. At the same time, however, we are working very closely with the major players in the industry, both in terms of technology and consulting, in order to make this phase productive. After all, despite cost pressure, a company must remain innovative and creative in order to remain competitive.
Artificial intelligence plays a key role here: where can AI be used in a targeted manner to create efficiency and competitive advantages? And where can additional savings be achieved through the use of modern software tools, both in terms of infrastructure and operations?
We offer very effective solutions such as Turbonomic, Instana and Apptio, which enable companies, including those in the automotive industry, to significantly reduce their IT costs while maintaining performance and quality.
Especially in difficult economic times, it is crucial to stand shoulder to shoulder and find ways to shape the future together. The automotive industry is and will remain the driving force behind Germany as a business location and we see it as our task to support them in this.
Are you worried that Germany is falling behind technologically, precisely because of this economic setback? Or do you think the country is still on the right track?
Wendt: I think Germany is back on the right track. Even if we are currently experiencing a dip, I believe that politicians, especially the current federal government, are doing a lot of things right by investing heavily in infrastructure.
Today, infrastructure means much more than just roads and networks, but also promoting innovation — and a lot is happening in this area. We at IBM are actively involved in this, for example in cooperation with universities and Fraunhofer Institutes, particularly in quantum research. This is how we maintain and promote knowledge and technological progress in Germany.
Of course, we know that some technologies, as far as their origins are concerned, originate in the USA. But with IBM’s location in Germany, we are bringing innovation and technology to the country, which will be used and further developed here.
We are also big advocates of open source components. This helps the location enormously because it enables openness and adaptability. One example is Red Hat, one of our subsidiary organizations, which makes globally developed open source components ready for production and secure to use.
I believe that all of this strengthens Germany as a business location. And as far as the art of engineering and know-how are concerned, I still see this as being very strong in Germany. I am optimistic: with the joint efforts of politics, industry and research, we are on the right track. Bitkom, with whom we work closely, also gets to the heart of the matter and has already achieved a great deal.
You have already touched on the topic of “digital sovereignty.” Have you noticed that customers are now placing greater emphasis on this topic, for example in the form of on-premises solutions?
Wendt: Yes and no. I can clearly see that demand is increasing, particularly due to the geopolitical situation and political discussions worldwide. Many companies, but also public authorities, want to regain their data sovereignty, i.e. manage their own data and retain responsibility and control. The call for this has become much louder.
At the same time, there are many marketing slogans on the market. Numerous manufacturers are now claiming that all of their hardware is now AI-capable and that every product is sovereign. That’s not true, of course. Real sovereignty only starts with the air gap. The term “digital sovereignty” is now somewhat worn out. I prefer to talk about technological autonomy.
Because true sovereignty means that systems must be able to be operated on-premises, completely separated from the internet and external influence, while remaining open, transparent and trustworthy.
These are values that we at IBM attach great importance to. We have to be careful that the term “sovereign solution” does not just become a label. It’s about substance and responsibility, not just buzzwords.
A good example is our Air Gap solutions, for example for customers in the security sector: systems that are guaranteed to be decoupled from the internet and yet are AI-capable and modern. That is true technological autonomy.
And yes, the demand for such solutions is growing, both in business and in politics. It is now crucial that truly responsible and resilient solutions are offered and we are working intensively on this.
