A comprehensive Flutter-based Android application designed to assist stroke patients with recovery through a chatbot and health resources.
A web application to visualize walking routes in Aalborg, Denmark, using the A* algorithm and Mapbox for real-time mapping.
Create beautiful, personalized greeting cards with stickers and GIFs.
A website that finds food discounts in Aalborg City, Denmark, from Føtex, Salling, Netto, and Bilka.
Research and implementation of intuitive hand gesture-based locomotion methods for virtual reality environments, focusing on natural movement without controllers.
• Designed and spearheaded the MVP development of an alternative healthcare platform to help people discover and search for treatments.
• Built the platform (frontend) using React, Next.js (App Router, SSR), Tailwind CSS, ShadcnUI/Radix UI, Firebase, and PostgreSQL.
• Owned end-to-end feature development, from domain modeling and SQL schema design to REST endpoints (FastAPI/Pydantic) and frontend UI.
• Delivered core features: searchable practitioner overview, onboarding and profile creation flows, profile editing (with image/video support), and cookie consent handling.
• Developed and deployed REST APIs and cloud solutions on Azure with FastAPI, SQLAlchemy, and PostgreSQL.
• Led architectural decisions on authentication and data models, including JWT/session handling and Firebase Auth integration.
• Mentored an intern, providing guidance on frontend development, solution design, technical best practices, and professional growth.
Designed a robust Linux-based controller for charging stations on Raspberry Pi Compute Modules, incorporating automated test environments, secure remote access, OTA updates with Mender, and a FastAPI REST API for streamlined user interaction
Contributed to the UX and UI development of KMD Clubtimiser (when Clubtimiser still was a part of KMD), a Microsoft Dynamics 365-based CRM for sports clubs, optimizing the customer information dashboard for usability and enhancing user experience through HCI principles.
My Master's thesis focused on generating user preferences based on walking routes in Aalborg, Denmark. Using a combination of Machine Learning, specifically ResNet, I developed a model to recognize and analyze Street View images, achieving over 90% accuracy in image recognition.
To model optimal walking routes, I applied the A* algorithm in a node system, allowing for pathfinding that adapted to user-defined preferences. The results demonstrated successful identification of preferred walking routes, effectively combining image recognition with weighted pathfinding to meet user needs.
All of this was implemented in a FastAPI REST API, providing a user-friendly interface for route generation and customization, the user also had the opportunity see and walk the suggested route virtually in a street view experience.
In my bachelor thesis, titled Comfortable Hand Gestures for Movement in Virtual Reality, I conducted a hand-gesture elicitation study to identify suitable gestures for locomotion control within a virtual reality environment. This project aimed to improve VR movement gestures, specifically continuous and teleportation gestures, using an elicitation-based approach.
Key research areas included defining gesture types, applying Wobbrok's study method with Morris' modifications, and minimizing legacy bias in user-generated gestures.The study involved 30 participants in both the elicitation study and a post-test, utilizing a logging system and NASA-TLX survey to assess task load.
The results indicated that gestures derived from the elicitation study had a lower task load index than industry-standard gestures, leading to discussions on the optimal number of gestures before quality declines. Additionally, I implemented software that could recognize these gestures in real-time within the virtual reality environment.