nutricue-ai-site

NutriCue AI Technical Architecture

Last Updated: October 5, 2025
Status: Live!

A comprehensive overview of the technical architecture behind NutriCue AI’s AI-powered reminder system, featuring backend-proxy AI integration, Railway deployment, and cross-platform subscription management.

Production System Overview

NutriCue AI is a full-stack application with a React Native mobile frontend and Railway-deployed Node.js backend, featuring secure AI integration via backend proxy, PostgreSQL database, and RevenueCat subscription management. The system is optimized for reminder management with HTTPS-only communication and App Store compliance.

Backend Architecture

Core Framework

Node.js - JavaScript runtime environment
Express.js - Web application framework
TypeScript - Type safety for backend services
JWT Authentication - Secure user authentication and authorization

Security & Middleware

Helmet.js - Security headers and protection
CORS - Cross-origin resource sharing configuration
Express Rate Limiter - API rate limiting and abuse prevention
Express Validator - Input validation and sanitization
bcryptjs - Password hashing and security

API Architecture

RESTful Endpoints - /api/v1/ versioned API structure
Health Monitoring - System status and uptime checks
Authentication Routes - User registration, login, device auth
AI Processing Routes - Reminder parsing and image analysis
Reminder Management - CRUD operations for reminders

Technology Stack

Mobile Frontend

React Native - Cross-platform mobile development
Expo - Development platform and build toolchain
TypeScript - Type-safe development environment
Expo Router - File-based navigation system
React Native Paper - Material Design 3 components

State Management

Zustand - Modern state management with TypeScript support
Zustand Persist Middleware - Automatic AsyncStorage persistence
AsyncStorage - Local data persistence with retry logic
Custom Hooks - Encapsulated business logic

Production Infrastructure

Railway Deployment - Production hosting at https://nutricue-backend-production.up.railway.app
PostgreSQL Database - Production database with connection pooling and automatic backups
HTTPS/SSL - Automatic certificate management via Railway
Health Monitoring - /health, /ready, /live endpoints for production reliability
Auto-Scaling - Railway infrastructure handles traffic growth automatically

AI Integration (Backend Proxy)

OpenAI GPT-4o - Natural language processing via secure backend proxy
OpenAI GPT-4o Vision - Image analysis for text recognition and scanning
Backend Proxy Security - API keys secured in backend, never exposed to mobile app
Usage Tracking - PostgreSQL-based limits and subscription validation
Cost Control - Backend-managed API usage and rate limiting

Subscription Management

RevenueCat SDK - Cross-platform subscription management with real API keys
Freemium Model - 5 active reminders, 3 image scans, 15 AI interactions/month free
Premium Tier - Unlimited reminders & AI conversations, 20 image scans/month
Pricing - $5.99/month or $39.99/year for AI Premium
Feature Gating - Usage limits enforced at backend level

Notification System

Expo Notifications - Local push notifications with smart scheduling
Platform-specific channels - Android notification channels for reminders
Smart Scheduling - AI-powered reminder timing optimization
Permission Management - Smart onboarding flow with skip handling

User Onboarding

2-Screen Flow - Welcome screen and permissions setup
Smart Permission Handling - Tracks permission requests and handles skips
Generic Content - Works for any reminder use case
Theme Responsive - Full light/dark mode support

Architecture Patterns

Production Data Flow (Backend Proxy)

Mobile App → JWT Auth → Backend API → Usage Validation → OpenAI GPT-4o → PostgreSQL → Response
     ↓              ↓              ↓               ↓              ↓              ↓
RevenueCat → User Validation → Rate Limiting → AI Processing → Usage Tracking → Mobile App

Subscription Flow

App Launch → RevenueCat SDK → Subscription Service → Usage Tracking → Feature Gating
     ↓              ↓                    ↓                 ↓              ↓
Local State → Platform Store → Backend Validation → PostgreSQL → UI Components

Security & Compliance

HTTPS-Only Communication - All production traffic encrypted
API Key Security - OpenAI keys secured in Railway backend, never exposed to client
JWT Authentication - Production-grade device authentication
PostgreSQL with RLS - Row Level Security for data isolation
Privacy Compliant - General reminder focus, no medical claims or prescription handling

Graceful Degradation

Offline Queue - Store AI requests for later processing when backend unavailable
Manual Fallback - Direct users to traditional reminder creation
Clear Messaging - Inform users of limited functionality
No Blocking - Core reminder functionality never depends solely on AI services

Production Architecture Benefits

Scalability & Performance

Railway Auto-scaling - Backend infrastructure scales with user growth
PostgreSQL Optimization - Connection pooling and query optimization
Rate Limiting - Prevents abuse and controls costs
Health Monitoring - Production reliability with comprehensive endpoints

Monetization Integration

RevenueCat Handles Complexity - Cross-platform subscription management
Backend Validates Subscriptions - Usage limits enforced before AI requests
Real-time Tracking - PostgreSQL-based usage monitoring
Graceful Limit Handling - Seamless transition when limits reached

Scalability Considerations

Local-First Processing - Reduces server load and latency
Efficient API Usage - Smart caching and request optimization
Platform Store Integration - Native payment processing

Development Practices

Testing Strategy

Jest - Comprehensive testing framework for frontend and backend
Supertest - HTTP endpoint testing for backend API
TypeScript ESLint - Code quality and consistency
90%+ Coverage - Comprehensive test suite requirements
Mock Services - RevenueCat, OpenAI, and database mocking

Build Pipeline

EAS Build - Expo Application Services for production builds
Automated Versioning - Semantic versioning with build number increment
Multi-Environment Support - Development, preview, and production configurations

Performance Optimizations

Notification Efficiency

Batch Scheduling - Minimize system notification overhead
Smart Recurrence - Efficient handling of complex repeat patterns
Memory Management - Proper cleanup of scheduled notifications

AI Processing

Backend Proxy - Secure API key management and request routing
Request Optimization - Minimal data transmission to OpenAI
Context Management - Efficient conversation history handling
Fallback Mechanisms - Instant offline mode switching
Cost Control - Usage tracking and rate limiting