Skip to content

ARCHITECTURE.md

Latest commit

 

History

History
36 lines (27 loc) · 1.52 KB

ARCHITECTURE.md

File metadata and controls

36 lines (27 loc) · 1.52 KB
Raw

NutriSense AI - Architecture

This document describes the high-level architecture of the NutriSense AI application.

Overview

NutriSense AI is a full-stack web application designed to provide intelligent, AI-powered nutrition analysis and personalized meal planning. It leverages Google Gemini for image analysis (food recognition) and Anthropic Claude for detailed nutritional breakdown and recipe generation.

Technology Stack

  • Frontend: React + Vite
  • Styling: Tailwind CSS + Framer Motion
  • 3D Hub: Three.js (React Three Fiber) for an immersive hero experience
  • Auth: Google OAuth for secure tracking
  • Deployment: Google Cloud Run (Containerized)

Key Components

1. Immersive Hero (Three.js)

A stylized, glowing 3D apple/heart symbol that reacts to mouse movement and scroll, symbolizing health and technology.

2. AI Nutrition Analyzer

  • Image Input: User uploads/takes photos of meals.
  • Gemini Vision: Analyzes the image to identify food items and portions.
  • Claude Optimization: Provides structured data (macros, calories) and health tips based on identified items.

3. Smart Recipe Engine

Generates recipes based on nutritional goals or existing inventory.

Data Flow

  1. User interacts with the dashboard.
  2. Actions trigger API calls (simulated for the demo or integrated via serverless functions).
  3. UI updates fluidly with Framer Motion transitions.

Future Roadmap

  • Integration with Google Fit/Apple Health.
  • Multi-user sharing (family plans).
  • Barcode scanning integration.