Simplifying the Diabetic Experience through User-Centric Engineering. A Semester-Long Project for CSE 3231: Software Engineering Techniques at The Ohio State University.
GlucoTrackAI was conceived as a comprehensive solution to the "Cognitive Tax" of diabetes. For Type 1 and Type 2 insulin-dependent users, managing health is often a fragmented, high-friction experience. This all started as a Hackathon Project back at SteelHacks 2025 (Pittsburg, PA) where me and a couple of my friends (Joselyn Vasquez, Krista Bair, Anika Talyan) were able to interview Krista's family friend Alex who had Type 1 Diabetes, and her mom who was her caregiver growing up since Alex was diagnosed at a very young age. This project to me aims to bring alive this concept from a quick prototype using replit to a working and functional prototype; ultimately, creating a new way for those with Type 1 & Type 2 diabetes to manage their condition better!
Our goal was to design an all-in-one, affordable, and hardware-agnostic platform that bridges the gap between patients, data, and caregivers. We replace "math fatigue" and app-switching with a streamlined, AI-integrated command center.
- Status: ✅ Phase 1 (Documentation & Research) Complete
- Current Focus: 🏗️ Phase 2 (High-Fidelity Prototyping)
Managing diabetes currently requires juggling multiple hardware silos (CGMs, insulin pumps) and manual logs.
- The Unified Dashboard: Consolidating disparate data streams (Dexcom, Medtronic, etc.) into one cohesive UI.
- Caregiver Sync: Real-time monitoring and "Last Action" logs to reduce caregiver burnout and anxiety.
- Affordability: A software-first approach that turns any smartphone into a medical receiver, eliminating the need for expensive proprietary handhelds.
This repository documents our systematic approach to the Software Development Life Cycle:
Comprehensive user stories and functional requirements mapping the needs of both primary users (diabetics) and secondary users (caregivers).
High-level system architecture diagrams illustrating data flow from external Health APIs to our secure backend, prioritizing low latency for glucose alerts.
Persona development and low-fidelity wireframes designed for accessibility. Features include high-contrast modes and "one-tap" insulin logging to accommodate users during hypoglycemic episodes.
Strategic plans for unit and integration testing, specifically focusing on HIPAA-compliant data handling and encryption at rest.
We are currently transitioning from theoretical documentation to a functional visual model.
- Design Phase (Figma): Crafting a design system that minimizes "Time to Action"—ensuring users can log meals in under 5 seconds.
- Logic Phase (Claude + React): Utilizing AI-assisted development to build out the frontend components in Next.js and Tailwind CSS.
- Integration: Developing mock API hooks to simulate live CGM data streams and predictive trend analysis.
| Field | Details |
|---|---|
| Institution | The Ohio State University |
| Course | CSE 3232: Software Engineering Techniques |
| Semester | Spring 2026 |
| Role | Architect & Researcher |
| Team | Melvin Vasquez, Armita Fard, Ethan Vincent, Zerin Merin, Mason Rush |
The core engineering work is located in the /docs folder. This project demonstrates the ability to translate complex medical requirements into a structured, scalable software blueprint.
Looking for the prototype? Stay tuned for the
/prototypedirectory, coming Summer 2026.