NutriScan

Scan smarter, live healthier, get instant nutrition and glucose insights from any barcode, designed to support diabetes-friendly decisions.

Comment

Inspiration

Diabetes management is a huge area where tech solutions can make a tangible difference. We wanted to build a tool that makes this easier, faster, and smarter; especially for people on the go. Inspired by the real-world struggles of friends and family living with diabetes, we envisioned a simple scanner that could guide users with real-time nutrition and glucose predictions.

What it does

NutriScan is a full-stack web app that lets users scan food barcodes using a live camera. It retrieves the item's nutritional information and uses a glucose dataset to predict the user’s post-consumption blood sugar level. The app then gives a clear recommendation: Safe, Caution, or Avoid; all in a matter of seconds

How we built it

Our app uses simulated insulin data based on the user’s food choices to provide real-time recommendations on whether it's safe to eat. While we’re using synthetic data in this prototype, the concept is fully scalable with real-time data from devices like Dexcom or Libre.

Challenges we ran into

  • Getting three different servers (React, Node, Flask) to work together without port conflicts.
  • Installing and debugging pyzbar and zbar on Mac systems.
  • Syncing real-time video scanning with backend prediction flow
  • Handling inconsistent data returns from external APIs
  • Making the UI user-friendly while preserving functionality ## Accomplishments that we're proud of
  • Successfully integrated barcode scanning with live video
  • Created a working prediction model based on real glucose data
  • Made an end-to-end experience where a user can scan and get results in under 10 seconds
  • Worked collaboratively across frontend/backend and overcame tough technical blockers as a team ## What we learned
  • How to build and sync multi-port full-stack apps
  • How to work with external APIs and real-world datasets
  • Dealing with system-level dependencies in Python (like zbar) ## What's next for NutriScan Deploy the full system for public access with cloud-based camera support

Built With

Share this project:

Updates