Selected projects up top, then coursework, hackathons, and internships - the ones solid enough to share.
Active Forecasting blood glucose 15-60 minutes ahead for Type 1 Diabetes, trained on ~2.5 years of my child's CGM and insulin-pump data. An automated pipeline feeds a 45-column feature matrix; an LSTM beats classical baselines (1.28 mmol/L RMSE at +30 min), with Clarke Error Grid analysis showing where accuracy and clinical safety diverge.
Active A budgeting app for couples and households, built to replace the fragile spreadsheet my wife and I outgrew. Plan a shared budget, forecast months ahead, and let an LLM sort credit-card statements into categories - with a Swedish tax engine and everything sensitive encrypted on your own device. In private beta.
An AI platform that helps small and medium-sized businesses manage continuity and risk across their digital supply chains, built with a team at Hive & Five during my internship and funded by a government grant (MSB). My first sustained work in a real multi-developer codebase: code review, CI/CD, and automated unit + e2e tests.
Active A local-first, fully offline desktop vault for the small secrets developers scatter across text files - API keys, snippets, commands, notes. Per-record ChaCha20-Poly1305 encryption with Argon2id key derivation; the master password never leaves the machine, and Win+V clipboard-history exclusion keeps copied secrets contained.
Active A lightweight, offline-first Markdown editor for notes, docs, and everyday writing - live side-by-side preview, tabs, and Word import/export via a custom Markdown-to-DOCX renderer. Runs in the browser or as a ~10 MB native desktop app built on Rust and Tauri. No account, no cloud, nothing leaves your device.
Archived LLM-powered hospital staff diagnose synthetic patients through a step-by-step workflow, visualized as characters moving through a 2D pixel-art hospital.
Paused A RAG assistant that answers questions about open-source libraries, grounded in their official documentation.
Archived Predicts taxi fares from ride details plus live distance, traffic, and weather data - a GradientBoosting model behind a FastAPI and Streamlit app.
Archived A data pipeline and dashboard for World of Warcraft auction-house prices, built in a month-long hackathon.