AI-powered food and medicine label analyzer — built with Expo React Native.
Created by Masun
Point your camera at any food packaging, supplement bottle, or medicine label. Label Truth uses Gemini AI to instantly break down every ingredient with color-coded safety ratings, a health score out of 10, possible side effects, and warnings.
- Camera mode — take a live photo of any label
- Gallery mode — pick an existing photo from your phone
- Health score — animated gauge (0–10), color-coded green / yellow / red
- Ingredient breakdown — Safe / Moderate / Avoid for each ingredient
- Side effects & warnings — AI-extracted for the general population
- Medicine profile — drug name, purpose, and dosage warnings for medicines
| Layer | Tech |
|---|---|
| Mobile | Expo React Native (SDK 54) |
| Navigation | Expo Router (file-based) |
| Camera / Gallery | expo-image-picker |
| AI Analysis | Google Gemini 2.5 Flash (via backend) |
| Backend | Node.js + Express + Multer |
| Charts | react-native-svg (animated health gauge) |
label-truth/
├── app/
│ ├── _layout.tsx # Root layout (fonts, providers, stack nav)
│ ├── index.tsx # Main scan screen (camera / gallery / analyze)
│ └── results.tsx # Analysis results screen
├── components/
│ ├── ErrorBoundary.tsx
│ └── ErrorFallback.tsx
├── constants/
│ └── colors.ts # Dark navy theme colors
├── assets/
│ └── images/
│ ├── icon.png
│ └── splash-icon.png
├── app.json # Expo config (bundle IDs, permissions)
├── eas.json # EAS Build config (APK / AAB)
└── package.json
- Node.js 18+
- Expo Go app on your Android / iOS device
- A deployed backend (see Backend section below)
# 1. Clone the repo
git clone https://github.com/YOUR_USERNAME/label-truth.git
cd label-truth
# 2. Install dependencies
npm install
# 3. Set your backend domain
# Create a file called .env in the root:
echo "EXPO_PUBLIC_DOMAIN=your-backend-domain.com" > .env
# 4. Start Expo
npm start
# 5. Scan the QR code with Expo Go on your phoneThe app sends images to a backend API that calls Gemini AI.
Endpoint: POST https://YOUR_DOMAIN/api/analyze
Body: multipart/form-data with field image
Returns: JSON with productName, healthScore, verdict, ingredients, sideEffects, warnings, medicineInfo
You can deploy the included backend or build your own Express server.
Required environment variables on the backend:
API_KEY=your_api_key
Use EAS Build from Expo to generate an installable APK.
# 1. Install EAS CLI
npm install -g eas-cli
# 2. Log in to your Expo account (free)
eas login
# 3. Configure the project (first time only)
eas build:configure
# 4. Build APK (installable on any Android device)
eas build --platform android --profile preview
# 5. Download the APK from the link provided
# Install it on your Android phone directlyThe preview profile in eas.json is configured to output an .apk file (not a signed .aab), so you can install it directly without going through the Play Store.
eas build --platform ios --profile previewRequires an Apple Developer account ($99/year).
| File | What to change |
|---|---|
constants/colors.ts |
Theme colors |
app.json |
App name, bundle ID (com.masun.labeltruth) |
app/index.tsx |
Scan screen UI |
app/results.tsx |
Results screen UI |
Copyright (c) 2026 Masun Technology. All rights reserved. Contact ceo@masun.qzz.io for usage permissions.
Created by Masun