PORTFOLIO
My Projects
These are the projects I've built from scratch - some worked beautifully on the first try, others took way more iterations than I'd like to admit. Each one taught me something new, whether it was a technical skill, a design principle, or just how to debug at 2am. I'm genuinely proud of what I've made here.
Nova Luna Studio
Co-design platform for educational games
Tracking Triumph
Track your every move
BriefBot
Efficient cooperation integrating AI.
Way To Go
A more enviromentally enlightened consumer
Machine Learning
Sentiment analysis and training of a CNN
Nova Luna Studio
Website
Nova Luna Studio was designed during the spring of 2025 as a part of my bachelor project. Nova Luna Studio's design-team consisted of Hallvard Lygre Hetlelid, Remi André Nygård, Silje Garmannslund Eide and myself. The prototype was designed as a deliverance to SLATE; The Norwegian Centre for the Science of Learning & Technology.
Background
The Norwegian Ministry of Education has outlined several
ambitions for digital practice in primary and secondary
education. Two of these are:
The use of digital solutions and media should be
knowledge-based and support students' learning, motivation,
and overall learning environment.
Digital practice, competence, infrastructure, learning
environments, and digital solutions should be integrated into
the school's comprehensive plans.
In collaboration with SLATE, we developed a prototype that aims
to turn these theoretical ambitions into practical solutions.
Target Group
Employees at SLATE // Tech-savvy teachers // Developers
Demographics
Teachers/educators/developers
Goals
Efficient and meaningful co-design of educational games
Pain Points
Educational games often lean too heavily on curriculum and lack engagement, or focus too much on fun with little educational value. Students who struggle in school may feel overwhelmed by challenging gameplay, while high-achieving students may find the games too easy and unengaging.
Research & Insights
To better understand how we could design Nova Luna Studio, we conducted a co-design workshop led by Fredrik Breien, the customer and creator of eLuna(read more about Breien's eLuna here). Alongside observation, 18 qualitative interviews were conducted with teachers, 1 researcher, 1 pedagogue and 1 developer.
Design Process
Iterative design, Crazy 8's & Storyboarding
The development of Nova Luna Studio followed an iterative design process, where insights from interviews and user testing were continuously used to refine the prototype. Instead of aiming for a complete solution in a single attempt, the process focused on continuous improvement, much like the production principles of the Toyota Production System.
Crazy 8's is a fast paced sketching method where each participant sketched 8 distinct ideas i 8 minutes. This part of the design process encourages creative thinking and help avoid sticking to the first idea.
Wireframing & Prototyping
The prototype was developed through several iterations. From low-fidelity on paper, made in the beginning of the design process to test ideas, to high fidelity prototypes in fima. A coded version was also prototyped. The Figma version prototype was used to explore user interface design, simulate interactions, and communicate the structure of the workshop process. The coded prototype focused specifically on implementing key features of the map-based canvas component in Nova Luna Studio.
Visual Design
The visual design of Nova Luna Studio naturally draws inspiration from the colors of the eLuna framework by Fredrik Breien. The "building blocks" of the games are colored according to Breien's original palette, making them easy to distinguish by shape, color, and function. The buttons in the game-creation process are also color-coded to match their corresponding blocks, ensuring clarity in the design process. Overall, the experience of Nova Luna Studio is intended to be a safe and supportive space where errors are allowed, manageable, and part of the creative process.
User testing
Informal user testing sessions with early prototypes were conducted to gather feedback on usability and concept alignment. To gather relevant user test data, we reached out to the three target groups: educators, developers, and SLATE employees. We user tested the prototype om five fifth-year teacher students at the Western Norway University of Applied Sciences, which led to great discoveries on how the platform should be designed to be intuitive for yeachers. Then, we user tested the prototype on five third-year information science students at the University of Bergen and gathered useful insight from the perspective of a developer. Lastly, we user tested the final prototype on the project's main stakeholder, SLATE.
Final Solution
Nova Luna Studio supports five core features. The platform supports two creation pathways, one offering a fully digital workflow, and the other providing a pen & paper option that enables teachers to run workshops before formalizing their ideas in the digital tool. This dual pathway supports flexibility and meets teachers where they are, whether they prefer digital tools or traditional methods. A collaborative canvas using eLuna’s visual language, making it intuitive for teachers to map out stories, characters, and challengers. Exportable XML/JSON files that serve as exportable blueprints for the developers. A demo game that is updated when the developers push new code. This is to show progress and to foster testing in classrooms. An informative landing page where users can see live notifications on projects, information about the projects, and who the collaborators are. You can read more aboud the final solution here.
Read more about my bachelor project
You can read more about my analysis and why I think Nova Luna Studio offers a new way of developing classroom tools for teachers here.
Test the prototype here!
Tracking Triumph
Body Tracking Software
Tracking Triumph was designed during the spring of 2024 in a design team consisting of Helene Herstad Bjørgen, Margethe Liahjell, Ole Gåsvær, Silje Garmannslund Eide and myself.
The prototype
Tracking Triumph is designed for the sports-addicted that aspire to get the absolute very best out of their efforts, much like Cristiano Ronaldo or Kristian Blummenfeldt. It's also designed for the athlete struggling with short or long-term injuries wishing to make a safe and quick comeback, like Zlatan Ibrahimovič or Jonas Abrahamsen.
Target Group
Early adopter athletes or tech-interested exerciser
Demographics
Male/Female 18-64.
Goals
Overview over the potential of the body and follow-up on injuries.
Pain Points
The need for a large number of technological trackers needed to track the entire body.
Research & Insights
To better understand how we could design Tracking Triumph, we conducted a semi-structured interview with a Ph.D student working within health and rehabilitation technology. This provided us with an understanding of how fitness applications are currently utilized by atheletes and in the health sector.
Design process
Ideation, brainstorming & wireframes
The design process consisted of Idea generation, How might we..?, Personas, Storyboard, and Wireframes.
Visual Design
We made a colour hierarchy in our system adhering to the 60-30-10 rule with dark blue, orange, and white. We consistently applied these colour across the system to maintain uniformity and standards. We maintained the colours at approx. 60%, 30% and 10%, with the orange colour chosen to evoke familiarity to other fitness apps.
Final Solution
3D-Model and Dynamic Layout
The prototype features a 3D representation of the user, displayed on the main page. It replays their movements in real-time and alerts them if any issues occur during a workout.
Quality of sleep
Providing in-depth analysis of sleep stages, helping users understand the distributing of sleep stages.
Workout Recap
Offers detailed reviews and comparisons of their workouts, helping them track progress towards their fitness goals. It includes weekly, monthly, and yearly analytics, along with specific warnings and visual guides like videos and images, to enhance training efficiency and safety.
Warning and Recommendations
Tracking Triumph monitors the users workouts and provides visual feedback using the 3D model, offering tips and analytics based on your performance. Additionally, the system's sensors accurately track movements, alerting users to any actions that could lead to injuries.
Evaluation
User testing
In our user testing, we specifically recruited participant from out target group: early adopter athletes. The recruitment ensured that the feedback we collected was relevant and insightful, contributing to precise improvements in the prototype functionality. Additionally, to maintain consistency and depth in out evaluation process, we involved the same Ph.D student we had previously interviewed throughout the user-testing. Read more about Tracking Triumph here.
Test the prototype here!
BriefBot
This design project was conducted during the fall of 2024 in collaboration with one of Norway's largest media organizations. The design and research team consisted of Agnes Villmones Haug, Anya Røng Skatvedt, Arian Velázquez, Frida Ege Olsen and myself. During the project, we focused on challenges related to Computer Supported Cooperative Work (CSCW).
Target Group
Emloyees of the media organization.
Demographics
Male/Female worklife age.
Goals
Make the information flow in the media organization more efficient.
Pain Points
- Increaslingly more information to handle.
- Large data storages for archiving of old media (required).
- Large amounts of customers, in a lot of different demographics.
Research & Insights
Several semi-structured interviews were conducted, and a thematic analysis approach was employed to identify patterns and themes in the data. To facilitate organization, the data was compiled into a FigJam file, where individual points were recorded on digital post-it notes. The process started with initial coding, focusing on identifying keywords or emphasized keywords. The post-it notes data were then categorised into three broad categories: Communication, collaboration and AI. These categories were further divided into seven more specific themes: Experiences, Communication for Collaboration, Information overload, Slack, AI, Innovation and Workshop. The analysis reveals the media company’s primary challenges: information overload, poor communication, and organizational inefficiencies, which results in time wasted, messages missed, and reduced productivity. For instance, a leader of a department described receiving hundreds of emails daily, often resorting to deleting them to cope, leading to stress and missed deadlines. These issues highlighted a need for streamlined processes and tools. Additionally, the leader of the EAI-team expressed significant interest in utilizing AI to address these challenges, given its potential to optimize workflows and reduce mundane tasks.
Design process
Ideation, brainstorming & wireframes
Initially, brainstorming sessions were conducted, during which team members contributed ideas tat could serve as the foundation for the prototype. Following this, the "How might we..." method was applied, an exercise that reframes our problems as opportunities. Personas also played a significant role in identifying the target user for the prototype. Four personas were created to represent the diverse range of workers who could benefit from the prototype. A storyboard was created to illustrate how an imaginary user could interact with the prototype. We also created vast amounts of paper wireframes to decide on key features.
Final Solution
BriefBot is an AI-powered tool, designed to streamline work and communication by automating tasks. It addresses challenges such as information overload, poor email management, and ineffective communication, while being intuitive and compatible with existing workplace applications.
Functionalities
The home screen is fully customizable, and the registration process is designed to be comprehensive, ensuring the AI gains a deep understanding of the job role. This foundation allows the AI to deliver personalized morning briefs tailored to the employee's specific needs. The AI not only learns during the registration process but continues to evolve by interacting with the employee, adapting its priorities to effectively
Automated calendar
Users have access to a dynamic calendar that updates automatically based on their schedules, ensuring no meetings or deadlines are missed despite information overload.
Morning reports
The tool generates AI-powered morning reports summarizing tasks and deadlines, helping users stay organized and productive.
Application compatibality
The tool seamlessly integrates with popular workplace platforms like Slack, Mail, and other communication tools, ensuring a unified user experience. Figure 9 illustrates this where users can select the applications to include.
Email & Slack
With AI, the user can summarize long emails and Slack chats for quick comprehension, translate emails, and automatically categorize emails based on importance:
Accessibility
BriefBot is designed with accessibility in mind, adhering to many Web Content Accessibility Guidelines (WCAG) and universal usability standards, with ongoing improvements aimed at achieving full compliance. Accessibility options, accessible via the gear-shaped 14 settings button, include adjustable font sizes for better readability, voice commands for hands-free interaction, and a text-to-speech reader to assist users who prefer auditory content. The interface also features a colorblind mode to address color vision deficiencies and a dyslexic mode to enhance text readability for users with dyslexia. Additionally, users can switch between light and dark modes to suit their visual preferences.
Evaluation
User testing
To evaluate the usability of the prototype, user-testing was conducted via Maze, which is a tool for conducting rapid user testing to collect qualitative and quantitative insights. The user-testing provided both quantitative and qualitative data, such as time spent and feedback after each task. Read more about the evaluation of BriefBot here.
Test the prototype here!
Way To Go
This design project was conducted during the spring of 2023. The design and research team consisted of Benjamin Late, Remi André Nygård and myself. During the project, we focused on challenges related to the UN sustainable Development Goals.
Target Group
Yong adults.
Demographics
Male/Female from "gen Z" and "millennials".
Goals
Make it possible for the consumer to scan their products to see the products' entire value chain and impact.
Pain Points
- Hard for the customer to distinguish enviromentally friendly things from not-so-friendly things.
- No information about value chain for the customer.
Research & Insights
Semi-structured interviews were conducted to gather good data for the target group. From the interviews, it was clear that the age group was moderately or very conscious about the environment. Eight groups with 1–5 users participated in the workshop. This provided the design with valuable data from users who could discuss different solutions and design proposals. The design also received individual thoughts from participants who gave input independently, without influence from others. During the workshop, it became clear that the users wanted an app with easy access to the scanning function, a start page similar to that of the Shazam app, and a clean application with multiple features.
Design process
Low-fidelity prototyping
Final Solution
From the locked screen (here: iPhone), the user clicks into the application and is taken directly to a green 'start' button. By clicking the start button, the user is brought to the application's main function: recognition of a product or scanning of a QR code. After scanning, the product will appear with selection options, the three figures below, and the possibility to scan again if the product does not match the desired item.
Functionalities
From the locked screen (here: iPhone), the user clicks into the application and is taken directly to a green 'start' button. By clicking the start button, the user is brought to the application's main function: recognition of a product or scanning of a QR code. After scanning, the product will appear with selection options, the three figures below, and the possibility to scan again if the product does not match the desired item.
Option 1 takes the user to information about the product’s journey, where it has been, and the product’s climate footprint. Here, the user also has the option to register the product; in that case, the following screen will look the same as if the user had chosen Option 2. Option 2 registers the product, including information about its climate footprint, and adds the product to the user’s already registered product library. Option 3 allows the user to see greener alternatives.
If the user clicks on ‘your statistics’ or selects ‘stats’ in the menu bar, the prototype takes them into the application’s statistics center. Here, they can view interesting statistics, both monthly and yearly. The user is also shown their greenest product, where they can click on ‘read more ->’ and be directed to the product’s information page outside the application. Colors can be seen as a consistent element throughout the entire design.
If the user clicks on ‘CO2 compensation,’ the prototype takes them to a page where they can choose to add an organization, read about organizations, or use points/money to support organizations that protect wildlife and the environment.
If the user clicks on ‘Green alternatives,’ they will be presented with several options for choosing greener. Since local choices were of particular interest among several workshop participants, the user can click on ‘Local choices’ and access a map showing local green alternatives, here represented by green pins. The green pins were suggested during the workshop.
Accessibility
If the user clicks on ‘Green alternatives,’ they will be presented with several options for choosing greener. Since local choices were of particular interest among several workshop participants, the user can click on ‘Local choices’ and access a map showing local green alternatives, here represented by green pins. The green pins were suggested during the workshop. If the user clicks on ‘Profile’ in the menu bar, they are taken to their personal profile, where different climate-related topics can be shared. The user can also post and view others’ ‘stories’ here, functioning as a kind of small social medium for climate issues. The UN wheel can also be found in the profile, reminding the user that the application is intended to inform about and help achieve the UN’s climate goals. The wheel will contain news articles related to the themes shown in the images. This serves as a unique element that sets the application apart from others.
Reflection
The design of the WayToGo application is not particularly unique within the context of apps. For example, the Vipps app has a feature that scans bills, and many other apps can scan QR codes. Many applications also offer statistics, and the social aspect is not unique in the market. The grocery provider Oda has a system for showing the climate footprint of food products (NHO, 2021), but it differs from WayToGo because it only displays products purchased from Oda. Nevertheless, the combination of features in WayToGo is unique, and in this niche of applications, there are no direct equivalents. WayToGo as a product therefore has no direct competitors. The idea of using image recognition and logistics information to create an app that illustrates consumption and travel distance came from strong knowledge of logistics. Perhaps due to current technological and knowledge limitations, this is the only design offering this on the market. Solutions that may seem similar to WayToGo include Shazam, but even there the differences are substantial. Since Shazam has likely conducted extensive research on button placement and other elements, it was natural to draw inspiration from it—for example, the placement of the first option to scan a product or certain menu elements. At the same time, all content and purpose are so different that major adaptations were required. This was one of many reasons why the workshop proved valuable for us. Yo can read more about Way To Go here.
Test the prototype here!
Machine Learning
This project in Machine Learning has absolutely nothing at all to do with UX/IU or interaction design. This project was completed in the spring of 2025 in collaboration with Henrik Sæther.
Read more about the project here!
