Virtual AI Trainer & Nutritionist Web App 🧠💪🍎

🚀 Overview

This is a personalized web application designed to act as a virtual fitness trainer and nutritionist. Users log their daily activities, meals, and other health metrics. An AI analyzes this data to provide personalized insights, recommendations, and eventually, workout/meal plans.

This project includes a Python (FastAPI) backend and a React frontend.

---

✨ Core Features Implemented

• Secure Authentication:
  • Sign up with Email/Password.
  • Log in with Email/Password.
  • Continue with Google (OAuth 2.0).
  • Secure JWT-based session management.
• User Onboarding:
  • Collects user profile data (age, weight (kg), height (cm), goal, activity level) after first login.
• Comprehensive Data Logging:
  • Log daily workouts (title, details).
  • Log daily meals (title, details, calories, protein).
  • Log daily sleep (hours, quality, stress level).
  • Log current weight (kg).
• Interactive Dashboard & Navigation:
  • Multi-page structure using React Router (Home, Insights, Chat, History, Charts).
  • Responsive Navbar with hamburger menu for mobile.
  • Home Page: Displays all logging forms (Workout, Meal, Sleep, Weight).
  • History Page: Shows scrollable lists of logged workouts, meals, sleep, and weight entries with delete functionality.
  • Charts Page: Visualizes trends for Calories/Protein, Sleep Duration, and Weight Change using Chart.js.
• AI Integration (Google Gemini):
  • Insights: Generates personalized insights based on user profile and recent logs on the Insights page. Includes insight history display.
  • Chatbot: Provides a conversational interface on the Chat page, using user data and chat history for contextual responses.
• Data Persistence & Logging:
  • SQLite Database: Stores user accounts, profiles, all log entries, and chat history.
  • Activity Logging: Records key user actions (signup, login) in the activity_logs table.
  • AI Interaction Logging: Saves AI insight prompts/responses in the ai_prompt_logs table. Chat history is stored separately.

---

🛠️ Tech Stack & Rationale

• Backend:
  • Framework: FastAPI (Python) - High performance, async support, data validation, API docs.
  • Database: SQLite - Simple, file-based for development. Uses SQLAlchemy (ORM).
  • Authentication: Passlib (hashing), Python-JOSE (JWT), Authlib (Google OAuth), Starlette Sessions.
  • AI: Google Generative AI SDK (Vertex AI) - For interacting with Gemini models.
  • Web Server: Uvicorn - Standard ASGI server.
  • Environment: Python Virtual Environment (venv).
• Frontend:
  • Library: React - Component-based UI development.
  • Build Tool: Vite - Fast development server and build process.
  • Routing: react-router-dom - For SPA navigation.
  • Styling: Plain CSS - Using App.css and Navbar.css for custom styling.
  • Charting: react-chartjs-2 & chart.js - For data visualization.
  • Package Manager: npm.
• Development:
  • Editor: VS Code
  • Version Control: Git

---

⚙️ Setup and Running

Prerequisites:

• Python (3.10+ recommended)
• Node.js (LTS recommended) and npm
• Google Cloud Project with Vertex AI API enabled.
• Google OAuth 2.0 Credentials (Client ID & Secret) configured for a Web Application.

Configuration:

1. Backend Environment (.env file):
   • Create a .env file inside the backend folder.
   • Generate secure random strings for SESSION_SECRET_KEY and SECRET_KEY (e.g., openssl rand -hex 32).
   • Add your Google Cloud AI API Key and OAuth credentials.
   • The file should look like this:

     SESSION_SECRET_KEY=YOUR_32_BYTE_HEX_SESSION_SECRET_KEY_HERE
     GOOGLE_API_KEY=YOUR_GOOGLE_AI_API_KEY_HERE
     GOOGLE_CLIENT_ID=YOUR_GOOGLE_OAUTH_CLIENT_ID_HERE
     GOOGLE_CLIENT_SECRET=YOUR_GOOGLE_OAUTH_CLIENT_SECRET_HERE
     SECRET_KEY=YOUR_32_BYTE_HEX_JWT_SECRET_KEY_HERE

2. Backend CORS Configuration:
   • Open backend/main.py.
   • Ensure the allow_origins list in CORSMiddleware includes your frontend URL (e.g., "http://localhost:5173").
3. Google Cloud OAuth Configuration:
   • In your Google Cloud Console (APIs & Services -> Credentials -> Your OAuth Client ID).
   • Ensure http://localhost:5173 is under Authorized JavaScript origins.
   • Ensure http://localhost:8000/api/auth/google/callback is under Authorized redirect URIs.

Installation & Running:

You need two separate terminal windows.

Terminal 1 (Backend):

# Navigate to backend directory
cd backend

# Create venv (if needed)
# python3 -m venv venv

# Activate venv
# macOS/Linux: source venv/bin/activate
# Windows: .\\venv\\Scripts\\activate

# Install dependencies
# (Create requirements.txt first: pip freeze > requirements.txt)
pip install -r requirements.txt

# Start server
uvicorn main:app --reload