Nibble AI
An AI nutrition tracker on the App Store: Claude identifies the foods in each photo, and a USDA lookup supplies the verified nutrition values.
Download on the App Store- Role
- AI Architect
- Type
- Independent
- Category
- AI/ML
- Stack
- Claude AI · USDA nutrition data · Claude Code · Swift · SwiftUI · HealthKit · Vision API · iOS
Home screen widgets
On the App Store
01
Overview
Nibble AI is an iOS application that revolutionizes nutrition tracking by leveraging Claude AI to instantly analyze meals from photos. The app eliminates the tedious manual entry that plagues traditional calorie counters, making healthy eating effortless and engaging.
02
The Challenge
Traditional nutrition tracking apps require users to manually search and log every food item, leading to low adherence rates. Users needed a frictionless way to track their nutrition that could understand complex meals and provide accurate nutritional information without extensive manual input.
03
The Solution
I designed and built an AI-first nutrition tracker around a photo-to-nutrition pipeline. Claude identifies the foods in each photo, and a USDA lookup supplies the verified nutrition values, so the model does the recognition and the numbers come from a trusted source.
I implemented barcode scanning for packaged foods, personalized AI coaching for dietary guidance, HealthKit integration for holistic health tracking, gamified challenges to boost engagement, and 16 home screen widgets for at-a-glance nutrition data.
04
Results & Impact
- Successfully shipped to the App Store
- 16 home screen widgets for personalized tracking
- Photo-to-nutrition pipeline: Claude identifies foods, USDA data supplies the values
- Seamless HealthKit integration for comprehensive health data