My thesis first, then image analysis and machine learning projects, then the software I've built along the way.
15 projects
Academic
Bachelor’s thesis investigating how affine data augmentation (rotation, scaling, translation) improves the performance of MobileNetV3-Small in handwritten signature identification using the BHSig260 Hindi dataset.
Academic · Uppsala University
Classifying cells from oral brush samples as coming from cancer patients or healthy ones, with only 19 patients. We placed ninth, and it taught me not to trust a validation score.
Academic · Uppsala University
A 4-stage classical image-analysis pipeline (FFT denoising, HOG features, SVM) that reads digits out of noisy, stripe-patterned CAPTCHA images — no deep learning or OCR allowed, 92.2% test accuracy.
Academic · Blekinge Institute of Technology
A Python-based implementation of edge detection from first principles, highlighting manual convolution, grayscale transformation, and image boundary handling—built without relying on image processing libraries.
Personal
An AI-powered predictive maintenance tool built with Streamlit to estimate the Remaining Useful Life (RUL) of turbofan engines using NASA’s degradation dataset. Deployed on Azure for real-time industrial insights.
Personal
Implemented the K-Means clustering algorithm in C for performance optimization, with Python scripts for testing, visualization, and data handling.
Academic · Blekinge Institute of Technology
Simulated the spread of COVID-19 across countries and age groups using a Markov chain model, visualizing infection dynamics through demographic-based data analysis.
Freelance
CropNeeds is a personal project focused on building a scalable e-commerce platform for farmers to browse and purchase agricultural goods using React, Firebase, and OpenAI API integration.
Academic · Blekinge Institute of Technology
CogniHive is a mobile app that blends educational collaboration with social networking, helping students find study partners, join study events, and enhance academic interaction.
Personal
Developed a lightweight Chrome extension that detects YouTube ads and temporarily increases playback speed to reduce ad viewing time without blocking or removing content.
Freelance
ReadMendations is a personal AI-powered book recommendation app that combines content-based and collaborative filtering, featuring synthetic reviews via GPT-3.5 and real-time REST APIs using FastAPI.
Academic · Blekinge Institute of Technology
Developed and deployed a Flask-based web application on an AWS EC2 instance with Ubuntu, demonstrating auto-scaling for enhanced reliability and performance in cloud environments.
Personal
Built a simple arithmetic expression evaluator in Python to explore lexical analysis, parsing, and interpretation—demonstrating foundational principles of programming language design.
Personal
Developed a cloud-native image processing system using Go, Docker, and Kubernetes. The service supports file uploads, image transformations, and microservice-based deployment via Minikube.
Personal
Built a desktop video player application using C++, OpenCV, and Qt, featuring playback functionality and brightness control. Serves as a foundational project for multimedia programming.
