Computer Science student and technical product engineer building practical software across desktop, spatial, AI, and web platforms.
I pair computer-science fundamentals with hands-on product engineering: defining the problem, designing the user path, building the system, and testing the result. My public coursework includes Java and Python projects from Bristol Community College and UMass Boston; outside class, I build products through Defendre Solutions.
FreeVoiceTranscribe — working, open source
Private, local hold-to-talk dictation for Apple Silicon Macs. The Python application combines global hotkeys, macOS Accessibility APIs, local Whisper inference through MLX, privacy-conscious storage, packaging, tests, and release automation.
Traces — active prototype, open source
A local-first knowledge workspace with an Electron and Next.js desktop shell, Markdown editing, a 3D knowledge graph, file-system tooling, and a multi-provider AI assistant. Built with TypeScript, React Three Fiber, CodeMirror, and Zustand.
LingBot Spatial — working visionOS demo, open source
Streaming 3D reconstruction concepts brought into Apple Vision Pro with SwiftUI and RealityKit. The project renders point clouds, camera paths, keyframe frustums, volumetric previews, and an immersive walkthrough from demo or exported reconstruction data.
A personal-finance dashboard exploring statement ingestion, transaction workflows, analytics, authentication, and a reactive Convex data layer. The repository remains private while the current data migration and import flow are completed, so I do not link to inaccessible source.
Airline Flight Schedule — academic Java project
A command-line scheduling system that models airlines, flights, departures, and arrivals. Recent improvements add JUnit coverage, safer input handling, and constant-time airline lookup with a LinkedHashMap.
- Personal portfolio — technical products, selected client work, and contact details
- Atomic Nature of Matter — Python coursework using depth-first search, image luminance, connected components, and particle tracking
- All public repositories
- Computer Science studies and applied Java/Python coursework
- Local-first AI and desktop software
- TypeScript product engineering with Next.js and Electron
- Swift and RealityKit experiments for spatial computing
- Testing, release automation, and maintainable delivery
The best public contact is [email protected]. You can also review my work at steve-defendre-portfolio.vercel.app.