GitHub - tung-programming/nutrigo

Smart Packaged Food Scanning Made Simple

Stop guessing what's in your packaged foods. Instantly scan packaged products with AI-powered precision to reveal hidden sugars, calories, and ingredients. Make informed choices in seconds.

🚀 Live Demo · 🐛 Report Bug · ✨ Request Feature

🎯 About the Project

NutriGo is India's AI-powered nutrition companion, revolutionizing how people understand packaged foods. In a world where food labels confuse more than they clarify, we bring clarity through intelligent technology.

Our Mission

To democratize nutrition knowledge through AI-powered technology, empowering every Indian to make informed food choices that improve their health and well-being.

Our Vision

A future where food transparency is the norm. Where every Indian has instant access to clear, reliable nutrition information that helps them live healthier lives.

Why NutriGo?

101 million Indians living with diabetes
Rising childhood obesity rates across the country
Confusing food labels that lack transparency
Need for instant, reliable nutrition insights

✨ Features

🔍 AI-Powered Scanner

Instantly scan any packaged food product with advanced AI to decode:

Scan chips, biscuits, breakfast cereals, snacks, and beverages
Identify sugar levels and hidden additives in processed foods
Get calorie breakdown and complete nutritional values
Analyze all ingredients and detect allergens
Works with Indian packaged brands like Lay's, ITC, Britannia, Amul, and more

🎯 Smart Health Score

Every packaged product gets an intelligent Health Score (0-100):

Comprehensive analysis of sugar content in snacks and beverages
Calorie density assessment for portion planning
Additive and preservative detection with warnings
Overall healthiness rating for instant comparison
Perfect for comparing products like Maggi, Haldiram's, Cadbury, and local brands

🌱 Better Alternatives

Discover healthier packaged food substitutes instantly:

Find healthier alternatives for favorite snacks and breakfast items
Compare popular products: Lay's vs Baked Chips, Regular Cookies vs Oats Cookies
See nutritional differences at a glance
Browse curated healthy options from brands like Yoga Bar, Slurrp Farm, Nature's Path
Integration with major e-commerce platforms (Blinkit, Zepto, Swiggy, BigBasket)

📈 Progress Tracking

Monitor your nutrition journey with detailed analytics:

Track your packaged food consumption patterns
Personalized recommendations based on scanning history
AI-driven insights from your dietary choices with real products
Weekly and monthly nutrition analytics for packaged foods
Visual progress charts and statistics showing improvement

⚡ Instant Analysis

Get real-time nutrition breakdown in milliseconds:

Scan packaged products and get results instantly
Lightning-fast AI processing of food labels
Complex nutritional data analyzed in seconds
Immediate actionable insights for better choices
No waiting, instant results for busy lifestyles

🤖 Smart AI Chatbot

Ask questions about packaged foods, nutrition, and ingredients:

Ask "Is Maggi noodles healthy?" - Get instant analysis
"What's in this snack?" - AI decodes all ingredients
24/7 instant AI-powered answers and recommendations
Voice input support (speech-to-text) for hands-free queries
Voice output support (text-to-speech) for audio responses
Conversational AI that understands your dietary goals

🎨 Beautiful Dashboard

Comprehensive user dashboard featuring:

See your recent packaged food scans at a glance
Nutrition statistics from all your scanned products
Health trends visualization based on packaged foods
Quick access to all scanning and comparison features
Responsive design that works on all devices

💾 Scan History

Keep track of everything you've scanned:

Complete history of all packaged food products scanned
Date, time, and details of each scan
Product names, brands (Britannia, ITC, Lay's, etc.), and health scores
Quick re-access to previous packaged food analyses
Easy search and filter capabilities

⭐ Favorites Management

Create your personal collection of packaged foods:

Save favorite healthy packaged products you like
Save unhealthy products to remember what to avoid
Quick access to frequently checked packaged items
Organize by category (snacks, cereals, beverages, etc.)
Track trending healthy packaged options

🔐 User Authentication

Secure and easy account management:

Sign up and login with email/password
Google OAuth one-click sign-in integration
User profile management (name, phone, location, bio)
Supabase Auth with session refresh middleware
Protected routes — middleware redirects unauthenticated users to login
JWT-based backend authentication with bcrypt password hashing
Privacy-first approach to your health data

🔄 Product Comparisons

Compare packaged products side-by-side:

Compare similar products: Britannia Cookies vs Sunfeast Cookies
View nutritional comparison charts instantly
See health score differences at a glance
Analyze ingredient variations between brands
Find the best packaged option for your dietary needs

📊 Analytics Dashboard

Advanced analytics for premium users:

Daily, weekly, monthly nutrition trends from packaged foods
Health goal progress tracking based on your scans
Detailed nutrient breakdowns (proteins, fats, carbs, sugars)
Personalized health insights and recommendations
Export reports of your nutrition journey

🎯 Health Goals

Set and track personal health objectives:

Customizable health goals
Goal progress visualization
AI recommendations aligned with goals
Success tracking

📱 Responsive Design

Seamless experience across all devices:

Mobile-optimized interface with touch-friendly controls
Tablet compatibility with adaptive layouts
Desktop full experience with sidebar navigation
Framer Motion animations and smooth transitions
Floating navbar with scroll-aware transparency

🌗 Dark/Light Theme

Full theme support powered by next-themes:

Dark mode by default with rich gradient backgrounds
Light mode support via theme toggle
System preference detection
Persistent theme preference across sessions

🎭 Animated Login Experience

Interactive and delightful authentication pages:

Animated mascot that follows your cursor on the login page
Mascot covers its eyes when you type your password
Happy expression on successful login
Starfield background animation with 60 animated stars
Google OAuth button with animated icon
Remember Me functionality

🍟 Supported Packaged Food Categories

NutriGo scans and analyzes products across all major Indian packaged food categories:

🍟 Snacks & Chips — Lay's, Bingo, Kurkure, Cheetos, Makhana, Popcorn
🍪 Biscuits & Cookies — Britannia, ITC, Sunfeast, Oreo, Parle-G
🥣 Breakfast Cereals — Kellogg's, Quaker Oats, Bagrry's, Muesli, Granola
🍜 Noodles & Pasta — Maggi, Yippee, Top Ramen, Pasta, Vermicelli
🧂 Instant Mixes — Dosa mix, Idli mix, Upma mix, Dhokla mix
🍫 Chocolates & Sweets — Cadbury, Nestlé, Amul, KitKat, Dairy Milk
🥛 Beverages — Tropicana, Real, Coca-Cola, Pepsi, Frooti, Paper Boat
🥛 Dairy — Amul, Mother Dairy, Paneer, Cheese, Yogurt, Ghee
🍯 Spreads & Jams — Peanut Butter, Nutella, Kissan, Honey
🍞 Bread — Sliced bread, Pav, Buns, Sandwich bread
🧊 Frozen Foods — Paratha, Samosa, Spring Roll, Nuggets
🥒 Pickles — Priya, Mother's Recipe, Patanjali
🥫 Condiments — Ketchup, Sauce, Mayonnaise, Chutney, Masala
💪 Health Drinks — Horlicks, Bournvita, Boost, Complan, Protinex
🥜 Nuts & Dry Fruits — Happilo, Farmley, Trail Mix, Roasted Makhana
🌾 Grains & Flour — Rice, Wheat, Atta, Quinoa, Ragi, Jowar
💊 Protein & Health Products — Yoga Bar, Slurrp Farm, Whey Protein

Each product is analyzed for optimal nutrition and health impact!

📥 PDF Export & Sharing

Download and share scan results and comparisons:

Export scan results as professional PDF reports (jsPDF + html2canvas)
Export side-by-side product comparisons as landscape PDFs
Share results via Web Share API or clipboard
Generate shareable comparison summary text
Dark-themed PDF reports with A4 formatting and multi-page support

🗣️ Voice Interaction (TTS & STT)

Full voice support powered by Sarvam AI and Web Speech API:

Text-to-Speech: Sarvam AI (bulbul:v1 model) with Indian English voice (vidya speaker)
Speech-to-Text: WebKit Speech Recognition for hands-free chatbot input
Browser-based TTS fallback when Sarvam API is unavailable
RecordRTC integration for audio recording capabilities
Real-time interim transcript display during voice input

🔬 Advanced Health Score Engine

Category-specific scoring with transparent breakdown:

8 specialized scoring algorithms: Beverage, Snack, Dairy, Packaged Food, Breakfast Cereal, Frozen Food, Condiment, Dessert
Detailed score breakdown showing penalties and bonuses with reasons
Score interpretation: Excellent (85+), Good (70+), Moderate (50+), Poor (30+), Very Poor (<30)
Confidence adjustment based on data completeness
Per-100g normalization for fair comparison

🤖 Multi-Layer Food Recognition Pipeline

Intelligent 4-step product identification:

Tesseract.js OCR — Extract barcode numbers and text from images
OpenFoodFacts API — Barcode lookup against global food database
OpenFoodFacts Text Search — Fallback search by product name
Gemini 2.5 Flash Vision — AI-powered image analysis with nutrition prediction when labels aren't visible

🏷️ Smart Category Detection

Automatic product categorization with subcategory support:

Detects 15+ categories: snacks, biscuits, beverages, sweets, dairy, grains, proteins, noodles, cereal, condiments, frozen, pickles, bread, spreads, health drinks, instant mixes
Subcategory detection (e.g., chips → chips/namkeen/popcorn, biscuits → cream/digestive/cookies)
Both frontend and backend implementations with regex-based pattern matching

🛒 Quick Commerce Integration

Direct purchase links to India's top delivery platforms:

Blinkit — 10-20 minute delivery links
Zepto — Ultra-fast delivery links
Swiggy Instamart — Quick grocery delivery links
BigBasket — Online grocery delivery links
Auto-generated search URLs for every alternative product

💰 Pricing & Subscription Tiers

Three-tier pricing model with yearly discounts:

NutriGo (Free) — 5 scans/day, basic health score, food comparisons
NutriPlus (₹249/mo) — Unlimited scans, detailed nutrients, AI insights, weekly reports, no ads
NutriPro (₹499/mo) — Everything in Plus + health goals, advanced analytics, hydration tracker, expert consultation
Monthly/yearly billing toggle with 17% yearly savings

🧠 CatBoost ML Model

Machine learning health score prediction:

CatBoost Regressor trained on Indian_Packaged_Food.csv dataset
Feature engineering: protein-to-calorie ratio, fat-to-calorie ratio, sugar-to-carb ratio, macronutrient balance, calorie density
RobustScaler normalization with 3-fold cross-validation
Model artifacts: health_score_catboost_model.pkl, health_score_scaler.pkl, feature_columns.pkl

📊 Dashboard Analytics

Rich data visualization with Recharts:

Weekly scan activity bar charts
Health score trend line charts
Streak tracking (consecutive days of scanning)
Healthy choices percentage calculation
Average health score aggregation

🍽️ Meal Tracking

Food record management by meal type:

Track scans by meal: breakfast, lunch, dinner, snack
Servings count per food record
Notes per scan entry
Aggregated average health score over configurable time periods (30-day default)
Indexed queries by user + meal type + date for fast lookups

🔔 Toast Notifications

Real-time feedback via Sonner toast system:

Success/error/info toast notifications across all actions
Scan saved, favorite toggled, PDF downloaded, profile updated confirmations
Auto-dismiss toasts with smooth animations
Global Toaster component on all dashboard pages

🧩 50+ UI Components

Comprehensive component library built on Radix UI + shadcn/ui:

Accordion, Alert Dialog, Avatar, Badge, Breadcrumb, Button, Calendar
Card, Carousel, Chart, Checkbox, Collapsible, Command Palette
Context Menu, Dialog, Drawer, Dropdown Menu, Form, Hover Card
Input, Input OTP, Label, Menubar, Navigation Menu, Pagination
Popover, Progress, Radio Group, Resizable Panels, Scroll Area
Select, Sheet, Sidebar, Skeleton, Slider, Switch, Table, Tabs
Textarea, Toast, Toggle, Tooltip, and more

🛠️ Tech Stack

Frontend

Next.js 16+ — React framework with App Router
TypeScript — Type-safe development
Tailwind CSS v4 — Utility-first styling
Radix UI — Full accessible component library (50+ components)
Recharts — Data visualization (bar/line charts)
React Hook Form + Zod — Form management with schema validation
Framer Motion — Smooth animations and transitions
jsPDF + html2canvas — PDF generation from DOM
Vercel Analytics — Production analytics
Sonner — Toast notification system
Lucide React — Icon library
date-fns — Date manipulation
cmdk — Command palette component
next-themes — Dark/Light theme system with system preference detection
RecordRTC — Audio recording capabilities for voice input

Backend

Express.js — Node.js web framework with TypeScript
Tesseract.js — OCR text recognition from food label images
Mongoose/MongoDB — Document database for products, users, food records, subscriptions
Supabase (PostgreSQL) — Primary database for scans, comparisons; Row Level Security
Winston — Structured logging with file transports (error.log, all.log)
Multer — File upload handling for image scans
express-validator — Request validation middleware
JWT + bcrypt — Authentication with JSON Web Tokens and password hashing
node-cron — Scheduled job infrastructure (cleanup jobs)
Morgan — HTTP request logging

AI & ML

Google Gemini 2.5 Flash — Vision AI for food label analysis and nutrition prediction
Google Gemini 1.5 Flash — Text-based food analysis fallback
Google Generative AI — Chatbot responses with knowledge base grounding
LangChain — AI framework integration (community + Google GenAI)
Sarvam AI — Indian English text-to-speech (bulbul:v1 model)
CatBoost — ML health score regression model (Python)
scikit-learn — Feature engineering, scaling, cross-validation

Data Sources

OpenFoodFacts API — Global food product database (barcode + text search)
Indian Packaged Food CSV — Training dataset for ML model
Knowledge Base FAQ — 25+ curated Q&As about Indian packaged food brands

Infrastructure

Vercel — Frontend deployment with edge middleware
Docker — Backend containerization (Node.js Alpine)
Docker Compose — Multi-service setup (app + MongoDB + Redis)
Supabase — PostgreSQL + Auth + Row Level Security

Authentication

Supabase Auth — Email/password + Google OAuth (via @supabase/auth-helpers-nextjs)
JWT (Backend) — Token-based API authentication with 7-day token expiry
bcryptjs — Password hashing with salt rounds
Middleware — Next.js edge middleware for route protection (/dashboard/*, /auth/*)
Row Level Security — Supabase RLS policies on comparisons table (users can only read/write their own data)

Testing

Jest — Unit testing framework
Supertest — HTTP assertion library for API tests
ts-jest — TypeScript Jest transformer

📂 Project Structure

nutrigo/
├── app/                        # Next.js App Router
│   ├── page.tsx                # Landing page (Hero, Features, About, CTA)
│   ├── layout.tsx              # Root layout with theme provider & chatbot
│   ├── globals.css             # Global styles
│   ├── api/                    # Next.js API routes (proxy to backend)
│   │   ├── auth/               # Login, signup, logout endpoints
│   │   ├── scan/image/         # Image upload proxy
│   │   ├── scans/              # CRUD for scan records
│   │   ├── alternatives/       # Alternatives proxy
│   │   ├── compare/            # Score recalculation proxy
│   │   ├── tts/sarvam/         # Sarvam AI text-to-speech
│   │   ├── metrics/            # Dashboard metrics
│   │   ├── health/             # Health check
│   │   └── user/profile/       # User profile CRUD
│   ├── auth/                   # Auth pages (login, signup, callback)
│   ├── dashboard/              # Protected dashboard routes
│   │   ├── scanner/            # Camera/upload scanner
│   │   ├── history/            # Scan history + [id] detail page
│   │   ├── alternatives/       # Healthier alternatives browser
│   │   ├── favorites/          # Saved favorites (localStorage)
│   │   ├── profile/            # Editable user profile
│   │   └── settings/           # App settings & preferences
│   └── pricing/                # Subscription pricing page
├── backend/                    # Express.js backend service
│   ├── src/
│   │   ├── app.ts              # Express app setup with CORS & routes
│   │   ├── server.ts           # Server entry point
│   │   ├── routes/             # API route handlers
│   │   ├── controllers/        # Request handlers (scan, auth, products, history)
│   │   ├── services/           # Business logic (chatbot, product, auth)
│   │   ├── models/             # Mongoose schemas (User, Product, FoodRecord, Subscription)
│   │   ├── middlewares/        # Auth, validation, error handling
│   │   ├── lib/                # External APIs (OpenFoodFacts, Supabase)
│   │   ├── utils/              # Health score calculator, category detector, logger, HTTP client
│   │   ├── knowledge-base/     # FAQ JSON for chatbot grounding
│   │   ├── config/             # Environment config
│   │   ├── jobs/               # Scheduled cleanup jobs (node-cron)
│   │   └── scripts/            # DB migrations & seed scripts
│   ├── Dockerfile              # Node.js 18 Alpine container
│   └── docker-compose.yml      # App + MongoDB + Redis stack
├── components/                 # Reusable React components
│   ├── scanner/                # Scan result, comparison, score breakdown components
│   ├── landing/                # Landing page sections (hero, features, about, nav, footer)
│   ├── ui/                     # 50+ shadcn/ui components (Radix-based)
│   ├── ChatbotWidget.tsx       # Floating AI chatbot with voice I/O
│   └── theme-provider.tsx      # Dark/Light theme provider (next-themes)
├── lib/                        # Frontend utilities & API clients
│   ├── api.ts                  # Auth, scans, comparisons API client
│   ├── health-score.ts         # Client-side health score calculation
│   ├── categoryDetector.ts     # 15+ category detection with subcategories
│   ├── comparisonUtils.ts      # Product comparison math & share text
│   ├── comparisonApi.ts        # Comparison CRUD API client
│   ├── comparisonContext.ts    # TypeScript types for comparison feature
│   ├── pdf-utils.ts            # PDF generation from DOM (jsPDF + html2canvas)
│   ├── getSmartAlternatives.ts # Smart alternative fetcher with match scoring
│   ├── mockAlternatives.ts     # 890+ lines of static healthy alternatives
│   ├── constants.ts            # App routes, health score ranges
│   └── supabaseClient.js       # Frontend Supabase client
├── model.py                    # CatBoost ML model training script
├── Indian_Packaged_Food.csv    # Training dataset for ML model
├── middleware.ts               # Next.js edge middleware (auth route protection)
└── hooks/                      # Custom React hooks (useToast, useMobile)

🏗️ Architecture & API Reference

Frontend API Routes (`app/api/`)

Route	Method	Description
`/api/auth/login`	POST	Mock login with deterministic user ID from email
`/api/auth/signup`	POST	Mock signup with email, password, name
`/api/auth/logout`	POST	Clear session
`/api/scan/image`	POST	Forward image upload to backend for OCR + AI analysis
`/api/scans`	GET	Fetch all scans for a user (by `userId` query param)
`/api/scans`	POST	Save a new scan record to Supabase
`/api/scans`	DELETE	Delete a scan (ownership verified)
`/api/scans/[id]`	GET	Fetch single scan detail by ID
`/api/scans/[id]`	DELETE	Delete specific scan
`/api/alternatives`	GET/POST	Proxy to backend alternatives API
`/api/compare/recalculate-scores`	POST	Forward score recalculation to backend
`/api/health`	GET	Health check endpoint
`/api/metrics`	GET	Dashboard metrics (total scans, avg score, streak)
`/api/tts/sarvam`	POST	Sarvam AI text-to-speech (Indian English, WAV format)
`/api/user/profile`	GET/PUT	User profile CRUD

Backend API Routes (`backend/src/routes/`)

Route	Method	Description
`/api/scan/image`	POST	Image upload → OCR → OpenFoodFacts → Gemini Vision pipeline
`/api/scan/barcode`	POST	Barcode lookup via OpenFoodFacts + Gemini fallback
`/api/scan/history`	GET	Fetch all scan history from Supabase
`/api/alternatives`	GET/POST	Smart alternatives with Gemini AI + static fallback DB (3000+ lines)
`/api/chatbot/chat`	POST	Knowledge-base grounded chatbot via Gemini
`/api/chatbot/health`	GET	Chatbot service health check
`/api/compare/save`	POST	Save comparison between two products
`/api/compare/history`	GET	User's comparison history
`/api/compare/:id`	GET/DELETE	Get or delete specific comparison
`/api/compare/recalculate-scores`	POST	Recalculate health scores with detailed breakdown
`/api/auth/register`	POST	User registration (MongoDB + bcrypt)
`/api/auth/login`	POST	JWT token-based login
`/api/auth/me`	GET	Protected route — get current user
`/api/products/alternatives`	GET	Healthier alternatives from Supabase (by minScore)
`/api/health`	GET	Backend health check with available routes

Frontend Pages

Route	Page	Description
`/`	Landing	Hero, Features, Market stats, About team, CTA, Footer
`/auth/login`	Login	Email/password + Google OAuth with animated mascot
`/auth/signup`	Signup	Registration with password confirmation + Google OAuth
`/auth/callback`	OAuth Callback	Handles Supabase OAuth redirect
`/pricing`	Pricing	3-tier subscription plans with monthly/yearly toggle
`/dashboard`	Dashboard Home	Stats cards, weekly charts, recent scans, streak tracker
`/dashboard/scanner`	Scanner	Camera capture + file upload → AI analysis → results
`/dashboard/history`	Scan History	All scans with search, filter by category, sort by date/score/calories/sugar
`/dashboard/history/[id]`	Scan Detail	Full nutrition breakdown, ingredients, warnings for a single scan
`/dashboard/alternatives`	Alternatives	Browse healthier alternatives with purchase links
`/dashboard/favorites`	Favorites	Saved favorite products (localStorage per user)
`/dashboard/profile`	Profile	Edit name, email, phone, location, bio; view subscription plan
`/dashboard/settings`	Settings	Notifications, dark mode, daily reminders, privacy controls

Backend Models (Mongoose)

Model	Fields	Purpose
User	name, email, password, subscription_tier	User accounts with bcrypt passwords
Product	barcode, name, brand, nutrition, ingredients_text, image_url, health_score	Cached product data from OpenFoodFacts
FoodRecord	user, product, scanDate, servings, mealType, notes, healthScore	Meal tracking with aggregation queries
Subscription	user, type (free/premium), status, startDate, endDate, paymentHistory, features	Subscription management with TTL index

Database (Supabase - PostgreSQL)

Table	Key Columns	Purpose
scans	id, user_id, product_name, detected_name, brand, category, barcode, health_score, nutrition (JSONB), ingredients, warnings, source	All scan records
comparisons	id, user_id, product_1_id, product_2_id, winner_id	Product comparison history with RLS policies
profiles	id, full_name, phone, location, bio, subscription_plan	User profile extensions

Key Scanner Components

Component	File	Description
`ScanResult`	`components/scanner/scan-result.tsx`	Full scan results with nutrition, score, alternatives, comparison, PDF download, share, favorite
`ProductComparisonView`	`components/scanner/ProductComparisonView.tsx`	Side-by-side comparison with VS badge, winner detection, PDF export
`HealthScoreComparison`	`components/scanner/HealthScoreComparison.tsx`	Health score visual comparison with score breakdown
`NutritionComparison`	`components/scanner/NutritionComparison.tsx`	Metric-by-metric nutrition diff with percentage indicators
`ScoreBreakdownCard`	`components/scanner/ScoreBreakdownCard.tsx`	Expandable penalties/bonuses breakdown card
`ComparisonCard`	`components/scanner/ComparisonCard.tsx`	Individual product card in comparison view
`ScanLoadingPortal`	`components/scanner/scan-loading-portal.tsx`	Full-screen loading portal during scan processing
`ChatbotWidget`	`components/ChatbotWidget.tsx`	Floating chatbot with voice I/O, Sarvam TTS, WebKit STT

Backend Services & Utilities

File	Purpose
`services/chatbot.service.ts`	Gemini 2.5 Flash chatbot with knowledge base loading from multiple paths
`services/product.service.ts`	Product fetch/cache: OpenFoodFacts → MongoDB with health score calculation
`utils/healthScoreCalculator.ts`	1028-line scoring engine with 8 category-specific algorithms + breakdown
`utils/categoryDetector.ts`	Regex-based food category + subcategory detection
`utils/httpClient.ts`	Axios client with request/response interceptors and logging
`utils/logger.ts`	Winston logger with console + file transports (error.log, all.log)
`lib/openFoodFacts.ts`	OpenFoodFacts API client (barcode lookup, text search, existence check)
`lib/supabase.ts`	Singleton Supabase client for backend
`middlewares/auth.middleware.ts`	JWT Bearer token verification
`middlewares/error.middleware.ts`	Global error handler with Mongoose/JWT error classification
`middlewares/validate.middleware.ts`	express-validator result handling

Frontend Libraries

File	Purpose
`lib/health-score.ts`	Client-side health score calculation with rating + color functions
`lib/categoryDetector.ts`	Extended 15+ category detection for frontend
`lib/comparisonUtils.ts`	Comparison math: diff calculation, winner detection, share text generation
`lib/comparisonContext.ts`	TypeScript types/interfaces for comparison feature + nutrition metrics config
`lib/comparisonApi.ts`	Frontend API client for save/get/delete comparisons
`lib/pdf-utils.ts`	PDF generation from DOM elements (A4, dark theme, multi-page)
`lib/getSmartAlternatives.ts`	Smart alternative fetcher with category detection, match scoring, fallback
`lib/mockAlternatives.ts`	890+ lines of static healthy alternatives across all categories with purchase links
`lib/api.ts`	Frontend API client for auth, scans, comparisons with safe JSON parsing
`lib/constants.ts`	App routes, health score ranges, nutrition guidelines
`lib/supabaseClient.js`	Frontend Supabase client instance
`middleware.ts`	Next.js edge middleware — session refresh + route protection for `/dashboard/*`
`components/theme-provider.tsx`	Dark/Light theme provider wrapping `next-themes`
`hooks/use-toast.ts`	Custom toast hook for Sonner notification system
`hooks/use-mobile.ts`	Responsive breakpoint detection hook

🚀 Getting Started

Prerequisites

Node.js 16+
npm or yarn
Docker (optional, for backend)
Google Cloud credentials
Supabase account

Installation

Clone the repository

git clone https://github.com/tung-programming/nutrigo.git
cd nutrigo

Install dependencies

# Frontend
npm install

# Backend
cd backend
npm install
cd ..

Set up environment variables

Create .env.local in the root directory:

NEXT_PUBLIC_SUPABASE_URL=your_supabase_url
NEXT_PUBLIC_SUPABASE_ANON_KEY=your_anon_key
SUPABASE_SERVICE_ROLE_KEY=your_service_role_key
BACKEND_URL=http://localhost:4000
NEXT_PUBLIC_BACKEND_URL=http://localhost:4000
SARVAM_API_KEY=your_sarvam_tts_api_key

Create backend/.env:

PORT=4000
GEMINI_API_KEY=your_google_gemini_api_key
SUPABASE_URL=your_supabase_url
SUPABASE_KEY=your_supabase_service_key
JWT_SECRET=your_jwt_secret
MONGO_URI=mongodb://localhost:27017/nutrigo
FRONTEND_URL=http://localhost:3000
NODE_ENV=development

Run the application

Backend (Terminal 1):
```
cd backend
npm run dev
```
Frontend (Terminal 2):
```
npm run dev
```
Open your browser Navigate to http://localhost:3000

💻 Usage

Basic Workflow

Sign Up/Login - Create an account or log in to access all features
Scan Products - Upload an image of any packaged food product
View Analysis - Get instant nutrition breakdown and health score
Track Progress - Monitor your nutrition journey over time
Find Alternatives - Discover healthier substitutes for your favorite products
Ask Chatbot - Get personalized nutrition advice 24/7

For Different Users

Health Conscious Users

Track daily nutrition intake
Set health goals
Get healthier alternative suggestions
Monitor progress with analytics

Busy Professionals

Quick product scans
Instant health scores
Fast alternative recommendations
No-fuss interface

Parents

Check kids' food safety
Identify allergens
Track family nutrition
Get health recommendations

Fitness Enthusiasts

Monitor macronutrients
Track protein intake
Compare supplement options
Optimize nutrition for goals

📱 Dashboard Features

Home Dashboard

Recent Scans - Quick access to last scanned packaged products
Nutrition Stats - Overview of your packaged food metrics
Health Trends - Visualize your nutrition journey from scans
Quick Actions - Scan new products, Compare brands, Ask Chatbot

Scan History

Complete list of all packaged food products scanned
Filter by date, category (snacks, cereals, beverages), or health score
Re-scan products you've analyzed before
See scanning patterns and top products

Favorites

Save favorite packaged products (both healthy and unhealthy)
Save Britannia biscuits, Lay's chips, Yoga Bar snacks, etc.
Organize by category and brand
Quick reference for common products

Profile

Manage account settings and health preferences
Update dietary restrictions (vegan, gluten-free, etc.)
View your subscription plan and usage
Download your complete scan history

Settings

Notification preferences for new features
Privacy controls for your scan data
Theme preferences (Dark/Light mode)
Language settings

Alternatives

Browse healthier packaged food options
Compare nutritional values of similar products
See purchase links to Blinkit, Zepto, Swiggy, BigBasket
Save alternative products to favorites

🤖 AI-Powered Tools

Smart Scanner Pipeline

Multi-step food recognition with intelligent fallbacks:

Tesseract.js OCR — Extracts text and barcode numbers from uploaded images
OpenFoodFacts Barcode Lookup — Searches global database by detected barcode
OpenFoodFacts Text Search — Falls back to text-based product search
Gemini 2.5 Flash Vision — Sends raw image to Google AI for product identification, nutrition extraction, and health assessment
Gemini Text Fallback — Text-based nutrition analysis when image analysis fails

Automatic category and subcategory detection
Nutrition prediction when labels aren't visible (with dataSource: "predicted" flag)
Scan result saved to Supabase with JSONB nutrition column

Health Score Algorithm (v3)

Category-aware scoring engine (1028 lines) with transparent breakdown:

Beverage scoring: Sugar/100ml penalties, calorie/100ml weights, protein bonuses for milk drinks
Snack scoring: Calorie, fat, sodium penalties; protein/fiber bonuses
Dairy scoring: Saturated fat focus, calcium/protein bonuses
Cereal scoring: Sugar-per-serving focus, whole grain/fiber bonuses
Dessert scoring: Aggressive sugar penalties, portion size awareness
Frozen/Condiment/General: Specialized algorithms per category
Score breakdown: Returns baseScore, finalScore, penalties[], bonuses[], summary
Score interpretation: Rating text + color for UI display
Confidence adjustment: Reduces reliability for incomplete product data

AI Chatbot Assistant

Knowledge-base grounded chatbot powered by Gemini:

Google Gemini 2.5 Flash model for response generation
Knowledge base: 25+ curated FAQ entries about Indian packaged food brands
Strict grounding: Only answers from knowledge base, declines unknown questions
Sarvam AI TTS: Indian English voice synthesis (WAV format, 24kHz, vidya speaker)
WebKit Speech Recognition: Browser-based speech-to-text for voice input
Fallback to browser SpeechSynthesis if Sarvam API fails
Available globally on every page as floating widget

Product Comparison Engine

Side-by-side product analysis:

Compare any two scanned products with inline scan-another-product flow
Health score recalculation via backend with detailed breakdown
Nutrition diff calculation with percentage differences and winner indicators
7 tracked metrics: Calories, Sugar, Protein, Fat, Carbs, Sodium, Fiber
Configurable lowerIsBetter per metric for accurate winner determination
Export comparison as landscape PDF or share via Web Share API / clipboard

Smart Alternatives Recommendation

Intelligent healthier product suggestions:

AI-powered: Gemini generates alternatives based on product category
Static fallback: 3000+ line database covering snacks, chips, biscuits, beverages, dairy, sweets, cereals, namkeen, noodles
Match scoring: 4 factors — health improvement (40%), subcategory match (30%), calorie similarity (20%), brand diversity (10%)
Purchase links: Auto-generated URLs for Blinkit, Zepto, Swiggy, BigBasket
Top 12 alternatives returned, sorted by match score

🐳 Docker & Deployment

Docker Setup

cd backend
docker-compose up -d

Services:

app — Node.js 18 Alpine backend (port 5000)
mongo — MongoDB 6 (port 27017, persistent volume)
redis — Redis Alpine (port 6379, persistent volume)

Database Migrations

Run in Supabase SQL editor:

-- Creates comparisons table with RLS policies, indexes, and updated_at trigger
-- See: backend/src/scripts/001-create-comparisons-table.sql

👥 Our Team

Meet the dedicated team of innovators building the future of food transparency:

Arjun Bhat
_{Frontend Developer}

Pranav Rao K
_{Frontend Developer}

Tushar P
_{Backend Developer}

Amogha K A
_{Backend Developer}

🌟 Key Highlights

✅ AI-Powered — Multi-model AI pipeline (Gemini Vision + OCR + OpenFoodFacts)
✅ Instant Results — Get nutrition insights in milliseconds
✅ Voice Enabled — Sarvam AI TTS + WebKit Speech Recognition
✅ Beautiful UI — 50+ Radix components, Framer Motion animations, dark theme
✅ Mobile Optimized — Responsive design with touch-friendly controls
✅ PDF Reports — Export scan results and comparisons as branded PDFs
✅ Secure — Supabase Auth, JWT, bcrypt, Row Level Security, edge middleware
✅ Free to Start — Three-tier pricing with generous free plan
✅ India-Focused — 17+ food categories, 25+ FAQ entries, 890+ static alternatives
✅ ML-Powered — CatBoost health score model trained on Indian food dataset
✅ Quick Commerce — Direct links to Blinkit, Zepto, Swiggy, BigBasket
✅ Dockerized — One-command backend deployment with MongoDB + Redis

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

📞 Contact & Support

Website: nutrigo-kappa.vercel.app
GitHub: tung-programming/nutrigo
Issues: Report a bug or request a feature

🌟 Made with ❤️ by the NutriGo Crew

Decode Your Packaged Foods, Redefine Your Health

Together, we're building a healthier India, one scan at a time. 🇮🇳

Name		Name	Last commit message	Last commit date
Latest commit History 159 Commits
app		app
backend		backend
catboost_info		catboost_info
components		components
data		data
hooks		hooks
lib		lib
public		public
styles		styles
.gitignore		.gitignore
Indian_Packaged_Food.csv		Indian_Packaged_Food.csv
LICENSE		LICENSE
README.md		README.md
components.json		components.json
health_score_scaler.pkl		health_score_scaler.pkl
middleware.ts		middleware.ts
model.py		model.py
next.config.mjs		next.config.mjs
package-lock.json		package-lock.json
package.json		package.json
postcss.config.mjs		postcss.config.mjs
tsconfig.json		tsconfig.json

License

tung-programming/nutrigo

Folders and files

Latest commit

History

Repository files navigation