About Me
Hi! I'm a 26-year-old Interaction Designer, newly graduated with a Master's degree and eager to apply my skills. I love exploring new technologies and leveraging them to create intuitive and impactful designs. When I'm not working or studying, you can find me wakeboarding - I even recently took home 3rd place at the Danish National Championship!
University Projects
Exploring conversational AI and response types for information elicitation
This project explores how conversational AI and choice-based responses can improve early-stage client-designer communication. This study, spanning three phases, focused on enhancing information gathering and fostering mutual understanding to achieve better design outcomes. We designed and evaluated an AI-powered tool to streamline client input, finding that while AI can prompt reflection and improve input clarity, it needs careful integration to avoid perceived slowness or restrictiveness. The system offers layered, editable AI outputs, including summaries and interpretations, which designers found valuable for understanding client needs. This work provides key insights for developing AI tools that support confident information sharing, allow for client-designer customization, and deliver synthesised, verifiable AI outputs to bridge communication gaps in design collaborations.
- Nuxt
- • Vue
- • Open AI API
Improving the sustainability of GenAI
This project explored how to minimize energy consumption in AI image generators while maximizing creative support. We evaluated the impact of prompt-to-image ratio and denoise level on energy use and creativity. In a study with 24 design and computer science students, we found partial denoising and single-image generation significantly reduced energy consumption. While these didn't impact creativity support, 20 participants preferred batching for exploration. Opinions on partial denoising were mixed. This research offers key design implications for sustainable generative AI.
- React
- • Stable Diffusion
- • Python
- • Local Image Generation
- • FastAPI
Zoodex: Collecting and Chatting with LLM-Animals
My research explores how Large Language Models (LLMs) and image classification can deepen visitor engagement and foster personal connections with animals in zoos. Through interviews at Aalborg Zoo and studies with university students, we identified a strong desire for interactive, educational experiences. This led to the development of Zoodex, an Android app designed for zoos. Zoodex uses image recognition, LLMs, and gamification to transform visitors into active researchers. Users "collect" virtual animals around the zoo, chat with unique AI personalities, and discover fascinating facts. An onsite evaluation with 20 pupils (12-14 years old) showed overwhelmingly positive responses. Participants loved collecting LLM-animals and found the interactive chats both immersive and educational. Zoodex not only facilitates learning and social interaction but also cultivates meaningful connections between visitors and animals through personalized AI interactions.
- Flutter
- • Dart
- • Open AI API
- • Image Classification
Retroscribe: Rewriting Classic Literature Into Contemporary Language
This project examines young adults' views on classic literature. A workshop with ten students aged 17-22, which consisted of focus group interviews, reverse brainstorming, and inspiration card, revealed that complex language in classics is a barrier. To address this, we developed Retroscribe, an AI-driven tablet app. It can rewrite classics into contemporary language while keeping the original plot. A hallway test showed a preference for the rewritten text, indicating increased excitement and relatability for the target audience.
- Open AI API
Generating routes to facilitate neighborhood exploration.
We created an application that generates routes based on user-selected points of interest (POI). We utilize various APIs, including OpenRouteService, a route generator built on OpenStreetMap, and Google Places, which retrieves images of POIs. After developing the first iteration, we conducted an evaluation to assess the overall user experience and get a System Usability Score (SUS). From this, we identified features for the final iteration of the application. Upon completing the second iteration, we conducted a comparative evaluation using the same methodology as the first evaluation.
- Unity
- • C#
- • OpenRouteService API
- • Google Maps API
Creating a motion controller for Overcooked!
Our objective was to create a controller for the game ”Overcooked! 2” that would enhance the gameplay experience through the incorporation of motion controls. Using 3D modelling and printing, and arduiono, we built a functional prototype. The performance of the prototypes was subsequently evaluated against a regular Xbox controller. The end result of our evaluation showed our controller made Overcooked! 2 more enjoyable through the use of motion controls.
- Arduino
- • Blender
Creating GuildMiners
This semester project focused on developing an interactive mobile game crafted with Unity, designed to connect people through shared interests and collaborative gameplay. Our methodology incorporated the Waterfall model, comprehensive user playtesting with questionnaires, and UI design guided by Affordances, Signifiers, and Gestalt Principles. The Octalysis Framework was leveraged to enhance player motivation. We successfully delivered a functional prototype, but we did not find the game to facilitate the theme of connecting people.
- Unity
- • C#
Personal Projects
Flush Pilot
An app for android where you can log a toilet visit and rate it from 1 to 5. Has social features like following friends. All data is saved in a relational database.
- Flutter
- • Dart
- • Supabase