How NN is redesigning customer processes with Agentic AI

AI-agents truly shine in such processes. Does that mean we shouldn’t deploy agents everywhere?

“Definitely. Some processes can easily be managed without AI—for example, if it’s simply a decision tree where no intelligence is needed. But where processes are more complex, AI-agents can play a huge role in our transformation and help elevate customer service to a higher level.

One example: imagine someone wants to apply for a mortgage at 11 PM after their evening shift. Thanks to AI-agents, we will soon be able to manage this process efficiently and user-friendly. The same applies to handling insurance claims or other customer inquiries. Internally, I often refer to the two-minute company: an organization where customers can manage their affairs in just a few minutes. Fast and convenient, whenever they want.”

“The rise of Large Language Models contributes to this. As a consumer, you experience daily how effective and user-friendly it is. That is exactly why applying this technology is not optional: financial institutions must respond to these expectations. It enables a complete redesign of how you serve customers and run backend processes. It won’t provide a lasting competitive advantage, at most, a first-mover advantage if, like us, you are the first to bring it to market.”

How do you engage employees in this transformation?

That is central to our approach because success is not determined by the technology but by how people interact with it. We have an AI literacy program for our entire organization of 16,000 employees, training them on what AI means. This is not only about using the technology but, more importantly, how it fits into NN’s transformation. We show what this means for NN’s strategy and, above all, what it means for each individual. People need to understand how they can contribute to this transformation. The principle of ‘lifelong learning’ has been in place for years, but with AI, it has really accelerated. Incidentally, as mentioned earlier, Alexander Klöpping’s media appearances have greatly raised awareness. We see this reflected in the strong interest in our training programs.”

Is there resistance because people fear AI-agents will replace their jobs?

“We tell it like it is: yes, AI replaces some human tasks, but it also creates new roles. We also show that work often becomes more interesting. Take automated claims processing: some cases are automatically approved or rejected, while more complex cases remain with humans. That is often the most enjoyable part of the job. Creative individuals now have more opportunities than ever. Processes are being redesigned in multidisciplinary teams; this is not just a task for AI specialists, but something done together with the business.”

You started exploring AI-agents relatively early. Can you share a learning experience?

“Not everything succeeds; you learn a lot and go through a steep learning curve. Setting up pilots is relatively easy, but the real challenge is scaling and bringing the solution into production.”

What would you tell organizations that are just getting started?

“Start with a clear ethical framework. Safe and responsible application is essential, and you simply cannot afford mistakes here. Because even in an era of constant change, one thing doesn’t change: trust comes on foot and leaves on horseback.”