Kantonsspital Winterthur

Alex Nelles

Alex Nelles is a customer-centric C-Level Executive with 25+ years of experience in Telco, Healthcare, Retail and Travel. As CIO at Kantonsspital Winterthur, he leads digital transformation, leveraging his fluency in "Business" and "Tech" to create strategic solutions that create value. Alex is passionate about enhancing quality and competitive advantage through technology and loves turning ideas into real-world advantages.

GenAI: Hype vs. Real-World Impact

I can't recall a tech wave that's created as much buzz as generative AI. The sheer scale of investment pouring into data centers, energy solutions and the pace of foundation model development is just mind blowing. It's like watching a rocket launch in real-time.

But there's always a 'but.' We need to keep our feet on the ground. Gartner's hype cycle for GenAI suggests we're heading for the 'Trough of Disillusionment' with LLMs and AI assistants in 2025 and that the 'Plateau of Productivity' might not arrive until 2026-2029. At the same time, we're seeing plenty of GenAI projects struggle, which is a natural part of any tech evolution.

On the other hand, pioneers are already proving GenAI's real-world value. We're seeing companies deploy chatbots that handle customer service with impressive autonomy, managing calls, scheduling appointments and answering emails. And let's not forget the explosion of AI-generated content in advertising, which is pushing creative boundaries and improving efficiency big time. Given the speed of innovation, I suspect commercial adoption may accelerate faster than many predict.

I see some genuinely transformative potential in healthcare, especially around the administration. Think about it: up to half of a clinician's time is spent on documentation. GenAI could revolutionize this by automating report generation or helping consolidate patient data into intuitive dashboards. Today, doctors must review several documents to understand the patient's past. Imagine doctors having instant access to a concise, AI-powered patient history overview.

The Slow Lane of Innovation in Healthcare

One thing that always makes me think is the imbalance of innovation in hospitals. The kind of technology being used in surgery or diagnostics is impressive and the machines are mind-blowing, but then you turn around and look at the administrative processes, and you see how far they are behind.

“GenAI won’t replace human expertise, but it can become a powerful assistant, helping staff focus on what truly matters: providing the best possible care for patients”

Part of the reason for this is that healthcare operates under a very different set of pressures compared to more commercialized industries. Most businesses are forced to reinvent themselves over time to survive, but hospitals are not managed that way. “If it isn’t broke, don’t fix it” is the motto for many processes.

AI Assistant Transforming Hospital Workflows

At Kantonsspital Winterthur (KSW), we’ve welcomed our first digital employee, Clara. And yes, she’s powered by GenAI!

Like many hospitals, we have many policies and guidelines and finding the correct information can feel like searching for a needle in a haystack. Clara simplifies this process by making key information instantly accessible. She’s already trained on many internal policies and regulations, providing precise answers while pointing directly to the relevant document sources. From now on, we simply ask Clara.

We started with HR, but this was just the beginning. Soon, Clara will become an expert in all areas of hospital operations, helping teams across departments navigate complex regulations and policies. While she’s still learning, we’re excited to see how she evolves and expands her capabilities.

Beyond internal policies, we’re also bringing GenAI to our patients. We’ve recently kicked off a project to build our patient portal and one of its key features will be an AI-powered chatbot. Before coming to the hospital, patients often have countless questions—where to go, what to bring or what to expect during their visit. Our chatbot will provide clear, instant answers to administrative queries and, if needed, connect patients with clinicians for medical-related inquiries.

Looking ahead, I see AI-driven agents becoming integral to hospital operations. These agents could assist staff with a variety of tasks - from registering patients for specific exams to drafting reports and supporting various steps in the treatment process. The potential is excellent and we’re just getting started.

GenAI in Hospitals: Between Promise and Panic

As we introduce GenAI into our hospital, we’re facing a mix of reactions from our staff. Some believe implementing AI should be straightforward since they already know tools like ChatGPT or Gemini. For them, AI is just another app - easy to integrate and ready to use.

What’s often overlooked is the complexity of embedding AI safely into our existing hospital environment. Unlike consumer AI tools, hospital-grade AI must interact seamlessly with legacy IT infrastructure while ensuring data security, compliance and reliability.

On the other end of the spectrum, there’s a group that struggles to see GenAI's potential in our operations. For them, AI feels abstract, more like science fiction than a practical tool that could make their daily work easier. They naturally hesitate to believe in it without a clear understanding of its capabilities.

And then, there’s another group - those who already recognize just how much AI is capable of. But instead of being excited, they feel fear. They see AI taking over tasks once done manually and worry about what that means for their own roles. These concerns are valid and that’s why it’s crucial to ensure AI is used to support, not replace, our workforce.

From Trials to Trust

A few key things need to change for GenAI to truly move from pilot projects to becoming a core part of hospital operations.

First, we need a shift in mindset. Right now, some see AI as just another tool, expecting it to work as quickly as ChatGPT, while others don’t see its potential at all. And then, of course, some worry about their jobs. To successfully integrate AI, we need to explain how it can streamline workflows, reduce administrative work and support staff rather than replace them. Clear communication, hands-on training and practical use cases will be essential to overcoming hesitation.

Second, we have to tackle the technical challenges. AI in a hospital isn’t just about answering questions - it has to work within our existing systems, ensure data security and comply with strict regulations. That means finding safe and efficient ways to integrate AI with legacy IT structures while protecting patient data. Without this, AI will remain stuck in isolated experiments rather than becoming part of our daily operations.

Finally, leadership buy-in and strategic investment are crucial. AI adoption won’t happen overnight and it requires commitment - not just in terms of budget but in promoting a culture of innovation.

Ultimately, GenAI won’t replace human expertise, but it can become a powerful assistant, helping staff focus on what truly matters: providing the best possible care for patients. Moving from experimentation to everyday use isn’t just about technology; it’s about people, processes and the willingness to be open to change.