Eduklar

Our tools cover various aspects of a teacher's daily tasks, with the goal of developing a service that can support as many teachers as possible, so they can spend more time on students and themselves.

We are a group of educators and web developers who have come together to simplify teachers' daily lives. The journey to Eduklar has been quick, educational, and fun. We have collaborated with teachers across Norway who have contributed through testing and discussions. These contributions have allowed us to offer tools that we believe can make a difference in the quality of teaching and contribute to sustainability for a profession often pressed for time.

There has long been speculation that AI technology will enable tailored teaching for each student, which could revolutionize education in the long run. I find it very exciting that this, to a limited extent, is already possible, which we have found through our previous studies in AI and text adaptation.

In August 2024, we launched Eduklar version 1 with limited functionality and a few tools. In October 2024, we updated to version 2 with a greater focus on student text adaptation, reporting, and more teacher-oriented functionality. The website is made with React, ExpressJS, MongoDB, Redis, is hosted on Railway and uses the newest and most powerful of OpenAI's models. This startup has taught me a lot about the process of developing a product, from idea to launch, and the importance of user feedback and testing, as well as the business and marketing aspects of a startup.

Studiehjelp

Studiehjelp began as an idea for helping students study more effectively. Together with a co-student, Trym, we further developed the idea into what it is today, a tool for teachers to adapt texts using AI technology based on students needs. It has been a long journey consisting of pitching contests, developing a startup business, development with new technologies, and doing research resulting in two conferencepapers.

The first conferencepaper was presented at the Seventeenth International Conference on Advances in Computer-Human Interactions in Barcelona. The paper entails interviews of 20 teachers regarding their routines when adapting education, as well as their perspectives about AI in the classroom. In this project, it's main purpose was to gaguge the viability and interest for an AI based text adaptation tool.

The second conferencepaper was published at the International Conference on Computers Helping People with Special Needs in Linz. This paper tested the efficacy of using AI for text adaptation by testing it against a teacher written text. We found that the AI adapted texts were rated to have a similarily high quality as the teacher written texts.

The website combines all our research into text adaptation with a genuine need for text adaptation tools into a userfriendly format. The website is made with React and Express.JS and uses the newest and most powerful of OpenAI's models to adapt texts automatically.

Collaborators:

• Trym Simensen Nerem
• Co founder

NTNUppl

NTNU pingpongleage is a website where users can participate in the schools ping pong league, keep track of their scores and compete for a spot on the leaderboard. Additionally matches can be played and scores can be updated, which will automatically count sets and winners. Finally users can play online by using the inbuilt websocket feature.

This project was a collaborative effort with two other web developers, and was the culmination of three previous projects being put together and improved. The website is built with a focus on user experience, with a clean and simple design, and a clear and intuitive layout. The website is also designed to be responsive and work on all devices.

In the final part of the project my main tasks were to implement the websocket feature, the email system, and parts of the authentication system and general backend. The authentication system uses refresh and accesstokens as well as email verification to ensure security. The websocket feature allows verified users to play online, in real time.

Collaborators:

• Trym Simensen Nerem
• Web Developer
• Ingrid Nes Gladheim
• Web Developer

Doom

This was the final project in our accessibility course. A solo project where there were no limits on the websites subjectmatter, so we could make something that appealed to our interests. I chose to make a fan site for the Doom franchise, with an emphasis on the newer games music. The site is made with HTML, CSS and JavaScript, and is designed to be accessible for people with disabilities, which was tested extensively with screen readers and keyboard navigation and user testing with visually impeaired users.

The website is designed to be easy to navigate with a screen reader, with a clear and consistent layout, and with a focus on the content which is primarily the lore and music of the Doom franchise. Additional focus was put on including aria labels, alt text and semantic HTML to make the site as accessible as possible.

Capmap

Capmap is a collaborative effort between me and another webdeveloper, and three interaction designers. The project consists of a website for users, a cloud hosted server, and a physical device consisting of a raspberry pi with an attached sensor. My role in this project was to do the electrical work and programming of the device, as well as making the frontend for the website.

The door mounted device is made out of a raspberry pi, and two ultrasonic sensors, all placed in a physical container made by one of the interaction designers. It uses GPIO pins to control the sensors, and a python script using MQTT to send information to the server. The device also has a small screen that displays it's ip address for easy setup.

Collaborators:

• Trym
• Web developer
• Aaron
• Interaction Designer
• Ruben
• Interaction Designer
• Henning Adolfsen
• Interaction Designer