Back to the career

Career

Clinomic

Senior Product Designer, Usability Lead · 08/2025 – present · Aachen

Rebuilding mona OS, Clinomic's critical care platform, alongside the Head of Product. Anything data-intensive in acute care can be modelled on it. Whether that becomes an anaesthesia information system, a patient data management system or an emergency department solution is a matter of configuration.

  • Medical device UX
  • Interaction design
  • Design systems
  • User research
  • AI-native product development
  • Information architecture
  • Object Oriented UX
  • Usability engineering
  • Design operations
  • Design leadership
  • Discovery and validation
  • Domain research
  • Data visualization
  • Agent-ready design systems
  • Design engineering

mona OS

  • Class IIa medical device

    Usability to IEC 62366-1

    As usability lead I own the documentation for certification. Every requirement can be traced from fieldwork into the product.

  • A prototype beside the product

    What a canvas cannot hold, I build as a working prototype beside the product. I write parts of the front end myself.

  • Tested before release

    Doctors and nurses test the flows formatively while they are designed. Representative users test them summatively before a release.

  • Safe by design

    Touch targets keep their size however dense the data gets. A person's entry outranks a device value.

  • Information architecture before screens

    Before I design screens, I map the core entities and how they relate, in collaboration with doctors and nurses.

  • Nothing averaged away

    On a patient's timeline a coarser view never averages. It shows the last real value and marks what it hides, so nothing relevant disappears.

  • A blueprint for medical interfaces

    I define the navigation and the floorplans for every view. Patient and allergies, for example, always sit in the same place.

  • From standard to interface

    I design the interface by the rules of FHIR. A wrong entry is not deleted but marked as an error. That way everyone sees what happened.

  • Peace of mind, not blanks

    Safe use starts with the users' mental models. A no has to be visible too: “No allergies” gives peace of mind, an empty field does not.

  • Friction that protects

    In medicine, friction also means safety. Where a wrong step can do harm, I build it in on purpose.

  • Loud only when it counts

    A screen gets loud only when the situation calls for it. Otherwise the real alarm drowns in the noise.

  • Complex tasks, manageable steps

    What looks like one click in a hospital is often a complex task. I break it down into manageable steps. That lowers the cognitive load, and nothing gets lost.

  • Designed for the real workplace

    Hospitals mean shared computers, gloves, switching between paper and screen, and constant interruptions. I design for that reality, not for someone in an office.

  • Patterns, not sprawl

    Many documentation tasks look different but run the same way. I work out shared patterns and make them available. New documentation tasks build on them instead of creating new variants every time.

Design system

  • 11 % → 76 % asked-for parts used

    AI-native design system

    Agents got only a basic sentence like “Build this screen” and no human help. On the first try they used 76 per cent of the required parts; without the design system, 11.

  • Contrast that is measured

    Colour has roles, and the contrast of every pairing is measured. Status never relies on colour alone.

  • Many standards, one design system

    A clinical product has to follow several standards at once. I bring them together in one design system. So every view is safe and easy to use from the start.

Role and team

  • Research in the hospital

    I have been in the operating room and the ICU and watched how the work is done there.

  • Design team

    I guide a small design team that keeps a much larger engineering team supplied with work it can build.

  • Directing AI agents

    Much of the prototype is built by AI agents. I write the specification, review at the irreversible step and catch what they miss.

  • Research beyond users

    Before a concept, I read what governs the work: clinical guidelines, professional agreements, safety notices and competing products, each claim labelled by how well it is sourced.

  • Critique before code

    Developers, doctors and nurses critique a design before anyone builds it. Their objections change it while it is still cheap to change.

  • Setting up design practice

    In my first weeks I set up the way design is done here: principles, critique format, personas, and usability engineering inside the regulated process.

  • Specs that cannot drift

    A design reaches engineering as a ticket with its reasons and testable behaviour, and its reference page is frozen so picture and requirement cannot drift apart.

  • A studio for design work

    I built a design studio that holds complex design work as data: open questions, decisions and the tests that settle them. Agents work in it over an MCP server.