Dr. Amigo 🩺

An elder's best friend - Your Personal Health Assistant

Project Overview

Dr. Amigo is a comprehensive health assistant application specifically designed for elderly care, with special focus on diabetes management and Alzheimer's support. This project provides two complementary versions:

• 📱 Mobile App (React Native/Expo) - Offline-first with local AI models
• 🌐 Web Application (Next.js) - Cloud-powered with full database integration

Both versions share the same mission: helping elderly people manage their health with dignity and independence.

🎯 Core Features

Health Management

• 🩸 Sugar Monitoring: Track glucose levels with color-coded warnings and comprehensive logging
• 📅 Health Calendar: Medical appointment scheduling and medication reminders
• 💊 Treatment Reminders: Insulin application alerts and prescription management

Memory & Family Support

• 👥 Family Photos: Photo recognition system for loved ones (Alzheimer's support)
• 📖 Memory Journal: Capture and preserve precious moments and experiences
• 🧠 AI Health Chat: Conversational interface for health questions and guidance

Technical Features

• 🔒 Privacy-First: Local AI models ensure data never leaves the device
• 🌍 Multilingual: Portuguese and English support
• ♿ Accessibility: Large fonts, intuitive UI designed for elderly users
• 🎤 Voice Interface: Speech-to-text for easy interaction

🏗️ Architecture

Mobile App (React Native + Expo)

📱 dr-amigo-mobile/
├── 🤖 Local AI Models
│   ├── Gemma 3n E2B IT (Chat)
│   └── Whisper Small (Speech-to-Text)
├── 📊 Features
│   ├── Sugar Monitoring
│   ├── Health Calendar  
│   ├── Family Photos
│   ├── Memory Journal
│   └── Voice Chat
└── 💾 Local Storage (MMKV + AsyncStorage)

Web Application (Next.js)

🌐 dr-amigo-web/
├── ☁️ Cloud AI (Together AI)
├── 🗄️ Database (Prisma + PostgreSQL)
├── 🔐 Authentication (NextAuth)
├── 📊 Same Feature Set
└── 🌍 Multi-tenant Support

🚀 Getting Started

Mobile App Setup

1. Clone and Install

git clone [repo-url]
cd dr-amigo
npm install

2. Install Dependencies

# iOS specific setup
npx pod-install ios

# Start development server
npm start

3. Run on Device

# iOS
npm run ios

# Android  
npm run android

# Web (fallback mode)
npm run web

4. Model Download
   • First run will prompt to download AI models
   • Gemma 3n E2B IT: ~4.2GB (main chat model)
   • Whisper Small: ~244MB (speech recognition)
   • Models stored locally for privacy and offline use

Web App Setup

1. Navigate to Web Directory

cd dr-amigo-web
npm install

2. Environment Setup

cp .env.example .env.local
# Configure database, auth, and API keys

3. Database Setup

npx prisma generate
npx prisma db push

4. Start Development

npm run dev

📱 Mobile App Features

Navigation

• Main Hub: Access all features from the enhanced conversation screen
• Chat Interface: Full-featured AI chat with reasoning mode
• Feature Modules: Dedicated screens for each health feature

Local AI Models

• Preset Configuration: No user model selection - optimized models pre-configured
• Offline Operation: Complete functionality without internet
• Privacy Guaranteed: All processing happens on-device

Data Storage

• Sugar Readings: Comprehensive glucose tracking with color-coded health indicators
• Calendar Events: Medical appointments with categorization and reminders
• Family Photos: Secure local storage with metadata
• Memories: Personal journal with emotion tracking and tags

🌐 Web Application Features

Multi-User Support

• User Authentication: Secure login system
• Personal Dashboards: Individual health profiles
• Family Sharing: Optional family member access

Cloud Integration

• Real-time Sync: Cross-device synchronization
• Backup & Recovery: Automatic data backup
• Advanced Analytics: Health trend analysis

Enhanced Features

• Context Injection: Daily health context for AI conversations
• Smart Reminders: Intelligent health alerts
• Progress Tracking: Long-term health monitoring

🎯 Target Audience

Primary Users

• Elderly Individuals (65+ years)
• Diabetes Patients (insulin-dependent)
• Alzheimer's Patients (early to moderate stages)
• Caregivers and Family Members

Use Cases

• Daily Health Management: Glucose monitoring, medication tracking
• Memory Support: Photo recognition, memory preservation
• Medical Organization: Appointment scheduling, health records
• Emergency Situations: Quick access to health information

🔒 Privacy & Security

Mobile App

• 100% Local Processing: No data transmission
• Device-Only Storage: All health data stays on device
• No Internet Required: Core functionality works offline

Web App

• Encrypted Storage: Database encryption at rest
• Secure Authentication: Industry-standard auth protocols
• GDPR Compliant: Full data privacy compliance
• Audit Trails: Complete activity logging

🌍 Internationalization

• Portuguese (pt-BR): Primary language for Brazilian SUS integration
• English (en-US): Secondary language for broader accessibility
• Extensible: Framework ready for additional languages

🏥 Healthcare Integration

SUS Compatibility (Brazil)

• Standard Formats: Compatible with Brazilian health record standards
• Export Features: Generate reports for healthcare providers
• Medication Database: Brazilian medication reference integration

Universal Design

• International Standards: Follows global healthcare data standards
• Customizable Units: Metric/Imperial system support
• Timezone Aware: Multi-timezone appointment scheduling

🛠️ Technical Stack

Mobile App

• Framework: React Native 0.76 + Expo 52
• AI Engine: Cactus React Native (local inference)
• Storage: MMKV + AsyncStorage
• UI: Tamagui design system
• Navigation: Expo Router v4

Web App

• Framework: Next.js 15 + React 19
• AI Service: Together AI API
• Database: Prisma + PostgreSQL
• Authentication: NextAuth.js
• UI: Tailwind CSS + Radix UI

📈 Development Roadmap

Phase 1 (Current)

• ✅ Core health features implementation
• ✅ Local AI model integration
• ✅ Basic family photo management
• ✅ Portuguese/English localization

Phase 2 (Next)

• 🔄 Advanced photo recognition
• 🔄 Medication database integration
• 🔄 Enhanced voice interactions
• 🔄 Emergency contact features

Phase 3 (Future)

• 📋 Healthcare provider integration
• 📋 Advanced health analytics
• 📋 Telemedicine features
• 📋 Wearable device connectivity

🤝 Contributing

This is an open-source project developed for the Google Gemma 3N Impact Challenge. We welcome contributions from:

• Healthcare Professionals: Medical expertise and validation
• Developers: Code contributions and feature development
• UX/UI Designers: Accessibility and elderly-friendly design
• Translators: Additional language support

Development Setup

# Clone repository
git clone [repo-url]

# Mobile development
cd dr-amigo && npm install

# Web development  
cd dr-amigo-web && npm install

📄 License

This project is 100% free and open source, dedicated to improving elderly healthcare accessibility worldwide.

🙏 Acknowledgments

• Brazilian SUS: Inspiration and requirements gathering
• Google Gemma 3N: AI model framework and challenge
• Healthcare Workers: Feedback and validation
• Elderly Community: User testing and insights
• Cactus Framework: Local AI inference capabilities

---

Dr. Amigo - Developed with ❤️ for the health and dignity of our elderly community.

"Technology should serve humanity, especially those who need it most."