Apps, packages, and experiments.
Tracks vocal tremors in Parkinson's patients using on-device machine learning. Won Apple's 2025 Swift Student Challenge (top 350 worldwide). Built for my grandfather.
My bet on what technology should be: tools that make people better at what they do, not tools that do it for them. As founder, I'm building a family of native Apple apps for people who make things. The software handles the busywork and disappears, so the human stays in the state that matters: thinking.
SwiftUI playground that turns camera roll photos into storybooks using Apple Foundation Models and on-device object detection.
Study companion that reduces digital distractions and builds healthy tech habits.
A self-hosted web app for taking drink orders when guests visit. They order from their phone, the host manages sessions from an admin portal. Built to learn Go, with a Swift CLI (torc) for one-command install on macOS and Linux.
Swift package for running LLMs locally on macOS. Privacy-focused, on-device inference with smart PDF extraction.
Filters JSON data by contextual relevance using on-device ML. Strips unnecessary fields to produce cleaner, more processable datasets. Part of the Tetrix Context Architecture.
SwiftUI package for Apple Pencil canvases. Document-style or infinite whiteboard.
JSON-based slide decks for SwiftUI. Themes, transitions, interactive elements. Built for AI agents.
Medication management for Parkinson's patients.