SELECTED WORK
Projects - Scroll w/ images or as list
-
01
GitHub →
MuZero Knockoff
From-scratch reimplementation of DeepMind's MuZero applied to 2048. Three neural networks trained jointly through backpropagation through time — the agent improves entirely through self-play without ever being told the rules.
-
02
Odelia 2025 →
Medical Image Analysis
Computer Vision final project using MONAI to analyze 3D medical scans for breast cancer detection. Compared DenseNet and ResNet architectures for the Odelia 2025 grand challenge.
-
03
GitHub →
Home Page
This website — a personal portfolio with CV and project showcase. Features a D3.js orthographic globe, scroll-driven journey section, full-screen project viewer, and dark/light theme toggle.
-
04
GitHub →
F1 DNF Analysis
Analysis and modelling of Did Not Finish outcomes in Formula 1 using historical race data. Tested SVM, Random Forest, Logistic Regression, and Neural Nets. Group project for MECENG278 at UC Berkeley.
-
05
GitHub →
Airport DB
Airport database simulation using a SQL backend and Python CLI interface. Built from a Relational Schema as part of TDT4145 Database Management Systems at NTNU.
-
06
GitHub →
PID-controller
JAX-based controller project comparing a classical PID-controller against a Neural Net controller across three plants: Bathtub, Thermostat, and Cournot market competition.
MuZero Knockoff
From-scratch reimplementation of DeepMind's MuZero applied to 2048. Three neural networks trained jointly through backpropagation through time — the agent improves entirely through self-play without ever being told the rules.
View on GitHub →Medical Image Analysis
Computer Vision final project using MONAI to analyze 3D medical scans for breast cancer detection. Compared DenseNet and ResNet architectures for the Odelia 2025 grand challenge.
View Odelia Challenge 2025 →Home Page
This website — a personal portfolio with CV and project showcase. Features a D3.js orthographic globe, scroll-driven journey section, full-screen project viewer, and dark/light theme toggle.
View on GitHub →F1 DNF Analysis
Analysis and modelling of Did Not Finish outcomes in Formula 1 using historical race data. Tested SVM, Random Forest, Logistic Regression, and Neural Nets. Group project for MECENG278 at UC Berkeley.
View on GitHub →Airport DB
Airport database simulation using a SQL backend and Python CLI interface. Built from a Relational Schema as part of TDT4145 Database Management Systems at NTNU.
View on GitHub →PID-controller
JAX-based controller project comparing a classical PID-controller against a Neural Net controller across three plants: Bathtub, Thermostat, and Cournot market competition.
View on GitHub →Beer Mile
A two-year longitudinal study in human resilience and determination. Team Lynrask's consecutive Beer Mile campaigns (2025–2026) demonstrate consistent improvement across key performance metrics and an unconditional commitment to crossing the finish line.
View results →
