Case studies

A selection of projects across physical prototyping and digital interfaces.

TimeTrace

Local-first AI journaling app that automatically tracks file edits and Git commits, generating audience-specific summaries with full privacy preserved.

Problem: Make developer progress clear, shareable, and easy to reflect on.
Outcome: Interactive calendar timeline with privacy-preserved summaries for PM, client, and self.

July 2025 – Present • AI • Privacy • Git

In Progress Read case

Subtext

AI‑powered prompt refinement tool that analyzes, refines, and rewrites what you type—making AI conversations clearer, more effective, and more human.

Problem: Most people don't know how to "talk to AI."
Outcome: Clearer prompts and better results inside ChatGPT.

Private Preview Read case

E-Ink Laptop

3D printed a case for a custom E-Ink laptop powered by a Raspberry Pi Zero and custom software to enable seamless note-taking with keyboard shortcuts and multi-file navigation.

Problem: Reduce screen fatigue and power usage while maintaining productivity.
Outcome: Custom laptop with E-Ink display, keyboard shortcuts, and multi-file navigation.

May 2025 – July 2025 • Hardware • 3D Printing • Raspberry Pi

Completed Read case Download GLB

Squirtle Sprayer

A playful, functional 3D-printed figure that actually sprays water—combining circuits, pneumatics, and delight.

Problem: Make a character come alive through motion and interaction.
Outcome: Robust mechanism with controlled spray and safe electronics.

Tools: Raspberry Pi, CAD, 3D printing, 12V pump, relays

Read case

GestureAR

Finger tracking and real-time overlays. Control LEDs based on hand position while augmenting faces.

Problem: Map expressive gestures to visible output.
Outcome: Reliable tracking with clear, legible AR treatments.

Tools: OpenCV, Python, Raspberry Pi, LEDs

GitHub

Hermes Lens

Wearable glasses that capture outside text and translate to your preferred language.

Problem: Translate on the go with minimal friction.
Outcome: Lightweight pipeline from OCR to on-device display.

Tools: Tesseract, OpenCV, Python, small LCD

GitHub

O.S.M.D

Scans and solves handwritten math problems using a Pi Camera and MathPix API.

Problem: Convert messy inputs into structured answers.
Outcome: Repeatable pipeline from capture to solution.

Tools: Raspberry Pi, Camera, LCD, MathPix API

GitHub