Stories from the work

Coach on Tap connects people with coaches: matching, discovery, demo sessions, and progress tracking. When I joined, the platform had already been redesigned twice. Both attempts had failed for the same reason: users still found it too difficult to use.

Working with the CEO and one UI designer, I spent three months rebuilding both the marketing website and the core platform, covering coach discovery, demo booking, and ongoing progress for coaches and mentees. I also defined the product approach for an AI feature, focusing first on where AI could genuinely improve the coaching experience and where it should stay out of the way.

The CEO was happy with the result as feedback from selected users was positive. One of the engineers on the team was particularly interested in the product approach I had taken. After the project, he reached out independently and asked me to join his company as a Product Owner.

It wasn't the first time work had led to that kind of conversation, but it still meant something. I took it as a pretty good sign that the way I had approached the product had landed differently.

Viacar AG provides IT services for road traffic authorities across Switzerland, including driver licenses, vehicle registration, exams, and more. Our job was to take their legacy desktop client, which operators had used daily for years, and rebuild it as a modern web application.

The tender phase happened before there was a contract to design for. I sketched user journeys and storyboards by hand, photographed them, and cleaned them up for presentation, whatever it took to help the sales team make the case to Viacar.

The harder problem came later. The legacy screens were dense, and operators had years of muscle memory tied to where information appeared. A generic modern form pattern would have made the interface harder to scan, not easier. Instead, the team developed something small: extending each label's baseline into a line running straight to its input field. No new visual element, just a stretch of something already there, and it solved the exact problem operators had lived with.

Modernizing the system did not mean replacing every old pattern with a newer one. For experienced users, familiarity was part of the usability. The better solution was to improve the interface without breaking the way people already knew how to work.

TL runs Lausanne's public transport network in Switzerland, including its buses and metro. At the time, a common ticketing transaction at the counter could take around seven minutes.

The bigger problem wasn't the number of screens. Staff were carrying too much of the interaction themselves: asking passengers for information, translating their answers into the system, and keeping track of decisions that passengers already understood better than they did.

We redesigned the flow around a touch device that let passengers enter the information they owned directly, while staff stayed focused on the parts that actually required their knowledge and assistance. That reduced both the workload and the cognitive burden on the person behind the counter.

For the common transaction we focused on, completion time dropped from around seven minutes to under thirty seconds.

What stayed with me was how willing people were to do a little more when the trade-off made sense. Entering the information themselves added a small step for passengers, but it took a lot of work off the staff and made the whole transaction much faster. That was reflected in the positive feedback the touch device received.