AI today is not just a technological topic. It is a change in thinking, way of working, and expectations – from customers and within teams.
At the last CX breakfast at ui42, we talked with experts from various fields about how AI is really entering CX and UX processes. The workshop was led by Barbora Blahušiaková, Peter Angelov, and Martin Krupa and brought a lot of practical experiences – from experiments to implementations in large companies.
What are CX breakfasts by ui42?
A closed discussion format where our colleagues host top CX experts from various companies and jointly discuss market challenges, customer issues, or technologies from the perspective of customer experience. Our regular guests include representatives from Tatra Banka, SLSP, SPP, ČSOB, O2, Telekom, Gardeon, MTBiker, Asseco, ....
AI integration starts with people, not strategy
AI in companies often doesn't start with big strategies. It only takes a colleague trying out new tools and naturally bringing them into the team. They become the internal "go-to" for AI and help find more efficient solutions, as expectations and demands for outputs are constantly growing. We all know today that correctly posed prompts are the foundation of correct answers. Therefore, more and more work is being done with pre-prepared prompts and guiding the user through the process. We can safely say that from a CX perspective, we are not only designing the result but also the interaction with AI itself.
AI as a partner in processes
AI is gradually shifting from the role of a tool to the role of a guide. For example, in one of the departments of the discussants, it helps business teams go through processes like discovery or problem definition.
However, users still do not perceive it as an authority but rather as a partner whose outputs need to be checked. This is why it is important to design solutions that account for verification.
At the same time, AI changes the dynamics of teams. Juniors produce quick outputs, seniors validate them. Work is not lost but rearranged, creating pressure for changes in education and mentoring.
Limits, experimentation, and adoption
AI has its limits - from hallucinations to internal barriers in companies, especially legal and security ones. This is why it often starts cautiously, for example, through local solutions.
On the other hand, AI significantly supports experimentation. People are less afraid to try new things because it doesn't cost them much time. Mistakes thus become a natural part of the process.
Customers also do not need to be pushed into AI. It doesn't have to be visible; on the contrary, it can work in the background and gradually improve the customer experience, similar to other innovations.
The discussion also revealed that AI only makes sense where it solves a real problem. And in some cases, a "good enough" solution may be better than waiting for perfection.
How we use AI at ui42
At ui42, we also approach AI practically. We have developed our own AI agent that helps us streamline work and minimize human error.
The most expensive mistakes are not the ones you know about, but those you don't – and that's where most "AI solutions" fail. They appear autonomous on the outside but actually work with third-party tools where there is no full control over data or the process.
We decided to go the opposite way and developed our own internal system that:
- works without third parties
- keeps data under our control
… and our clients do not have to worry about the leakage of sensitive information.
The workshop also clearly showed that the key to good outputs is not just the AI itself, but the ability to ask the right questions. This is where our expertise lies. We know how to work with AI, how to correctly "navigate" it, and how to get relevant and quality outputs from it. This is why companies that want to use AI effectively but don't know where to start turn to us.
AI is not just about tools. It's about changing the approach, culture, and way of working. And that's what our CX breakfasts are about – sharing experiences that move things forward.
