I like turning ambitious ideas into polished, usable products.
CS and Electrical Engineering student at Drexel University, driven by a constant need to build things that actually matter and understand them at a level deeper than surface. I thrive under pressure, most recently placing 1st overall at HackPrinceton Spring 2026, Healthcare Track. Hackathons aren't hobby events to me; they're the closest thing to a real proving ground where execution is the only currency. My work spans healthcare data systems, low-level simulations, and everything in between, always with the same underlying question: what does it mean to build something you can actually trust? I care about reliability and clarity far more than complexity, and I'm skeptical of any system, AI or otherwise, that can't explain itself. Philosophy runs alongside the technical work, not separate from it. Tesla put it well: "The present is theirs; the future, for which I really worked, is mine." That's roughly how I think about the problems I choose and the standards I hold myself to. Outside of building, I'm competing, reading, thinking, and trying to understand the world a little more precisely than I did yesterday.
- AI tools for messy real-world workflows
- Full-stack products with clean UX and strong product feel
- Research prototypes that can grow into real products
- Systems that automate repetitive work without unnecessary complexity
- Demos that are technically strong and easy to explain
Privacy-preserving mediation layer for clinical trial workflows using local models, Safe Harbor stripping, routing, and differential privacy.
Goal: Make LLM use safer in high-stakes clinical workflows
Focus: Privacy, utility, and product usability
Stack: Python, local model routing, experiments, paper, demo
MedTrack is a local-first medication adherence app built for hackathon demo use. It helps users stay on schedule with reminders, log dose outcomes, review adherence, and access safety-focused guidance.
What it does:
- Medication management with add/edit/delete/deactivate flows
- Flexible scheduling for fixed times, interval-based doses, day-of-week plans, and PRN usage
- Local reminder notifications with snooze support
- Dose action flow for Taken / Snooze / Skip
- Adherence history, filters, and status tracking
- Safety Check, Missed Dose Guidance, and Emergency Card support
- Demo mode for live walkthroughs
Goal: Build a polished, safety-conscious medication tracking experience
Focus: Strong UI/UX, local-first reliability, and clear patient workflow design
Stack: TypeScript, React Native, Expo, Expo Router, SQLite, Expo Notifications, Expo Camera
WeaveWise is a clothing sustainability tracker that reads garment tags, extracts structured clothing data, and estimates environmental impact for a single item or an entire wardrobe.
What it does:
- Extracts material composition, country of origin, and care details from garment tags
- Uses OCR + LLM parsing to turn messy label text into structured garment data
- Enriches results with sustainability context and impact factors
- Generates garment-level and wardrobe-level impact summaries
- Helps users understand the environmental footprint of their clothing choices
Goal: Make clothing sustainability information easier to access and understand
Focus: OCR, structured extraction, sustainability analysis, and wardrobe-level insight
Stack: React, Vite, TypeScript, Python, FastAPI, MongoDB Atlas, Groq API, LangGraph, Bright Data
- Building stronger hackathon demos
- Building end-to-end products with better polish
- Building software with real workflow value
- Learning deeper backend systems
- Learning practical ML/AI integration
- Improving architecture, UI presentation, and iteration speed
If you liked my work or if it has helped you, all support is appreciated!
