Flobak Lab
CurrentComputer vision for cancer research, helping identify suitable chemotherapy earlier.
Software & machine learning
Hi, I’m Theodor.
Full-stack developer and computer engineering student at NTNU. Working on applied machine learning through ReLU.
NTNU · Trondheim, Norway
About
I’m a computer engineering student at NTNU working in software development, with a particular interest in machine learning, product development, and the intersection of technology and design.
I believe the most interesting problems are solved by challenging established assumptions, rather than following consensus. I’m particularly interested in first-principles thinking, creating things that don’t exist yet, and finding opportunities where conventional wisdom may be wrong.
I have experience ranging from full-stack development and machine learning to interaction design and entrepreneurship. I enjoy moving between technical depth and the bigger picture: understanding how things work, why they matter, and whether we’re solving the right problem in the first place.
I’m a technology optimist, but skeptical of hype. In machine learning, I’m more interested in models, research, and fundamental principles than in the latest frameworks and abstractions.
I value independent thinking, thoughtful design, privacy, and collaboration with ambitious people who are willing to challenge each other’s ideas. Ultimately, I’d rather help create something new than make marginal improvements to something everyone already agrees on.
Student organisation
ReLU NTNU
Machine Learning Engineer · student organisation · Oct 2024 — Present
Through ReLU, I work on applied machine learning with industry and research partners alongside my studies at NTNU.
Schibsted Media
Fine-tuned vision-language models for internal use.
Triplan
Document classification with machine learning.
Recognition
Hackathons, competitions & awards.
From national entrepreneurship to hands-on machine learning. Building, competing, and learning together.
2023 — 2024
Norgesmester i Ungt Entreprenørskap
Norway’s best youth business · 2024Euronext Blue Innovation Award · 2024
My team and I won Norway’s national youth entrepreneurship competition, with around 2,500 companies and 14,176 pupils participating, and the international Euronext Blue Innovation Award. We developed a circular business model for infrastructure in low-trophic aquaculture.
The Euronext Blue Innovation Award brought together nine countries. Of 23 Norwegian businesses nominated, seven were selected for the Norwegian championship. We went on to become the single international winner across all nine countries.
Our work also involved marine community gardens, the campaign for a cleaner Oslofjord, beach clean-ups, and a youth environmental group. We shared experiences with European groups working to maintain marine habitats.
Hackathon
Astar Consulting & NorgesGruppen Computer Vision Hackathon
First place
Our team took first place after 17 hours of programming. We trained Ultralytics YOLO and ResNet models to track and classify self-checkout video, using balanced data augmentation and a voting system to improve detection when model confidence was low.
YOLOResNetObject trackingData augmentationVoting system
Our team: William M., Theodor Strøm-Thrane, Eldar Alvik, and Tobias Fremming
2025
IES & ELSYS Innovation Competition
Innovation prize
Developed a solution for fault detection in cooling tanks and won the innovation prize.
14–15 Mar 2025
DNV–BRAIN Hackathon
Third placeFunniest presentation
With team Crash Course, I built a YOLO v11 model to detect and classify maritime vessels during this 24-hour hackathon. We used manual labelling, data augmentation, and Grad-CAM to explain the model’s predictions. We placed third overall and won the funniest presentation prize.
YOLO v11Object detectionGrad-CAMData augmentation
2024
StartCode
Medal
A green-energy management system.
2023
TreasureHacks 3.5
Medal
Combining language models and flashcards to build a new way to learn.
Skriv for fremtiden · national writing competition
Won the United Nations Norway writing competition with a piece on efficient resource use. Spoke virtually on national television through the Kåkånomics festival.
Jobs & experience
Select a role to read moreMigrated authentication from Clerk to WorkOS across the backend, web, mobile app, and supporting infrastructure. Built parts of the app from scratch and improved core functionality and data models for product packaging and AI-powered document processing.
Full stackAuthenticationAI products
Studies
NTNU
2024 — PresentBSc Computer Engineering · Systems Development
Additional coursework in computer vision, deep learning, AI, control engineering, linear methods, and contract law.
Elvebakken videregående skole
2021 — 2024Information Technology & Media Production
University-preparatory studies, including Physics 1, business development, and R2 mathematics. Grade average: 5.6.
Tools & skills
Development
TypeScriptJavaScriptPythonJavaC++ReactNext.jsNode.jsReact NativeExpoSQLNoSQLNeonDockerGit
Machine learning
PyTorchComputer visionCNNsTransformersRAGFine-tuningU-NetsSelf-supervised learning
Design & media
FigmaPhotoshopIllustratorPremiere Pro
Earlier side projects
A few older experiments along the way.
- ChatCardsConversational flashcards built at TreasureHacks 3.5.
- Page PrankerAn experimental Chrome extension for browser pranks.
- PhotosolveA mobile app that scans questions and answers them with a language model.
Contact
Let’s build something.
Have an interesting problem, a project in mind, or just want to talk software and AI? I’d love to hear from you.
Send me an email