Transform your health journey with a full-stack AI-powered diet and nutrition assistant. This project leverages modern web technologies and local LLMs to deliver personalized meal plans, nutrition analysis, and smart recommendations, all in a beautiful, user-friendly interface.
AI Diet Assistant is a full-stack application that helps users:
- Generate personalized diet plans based on their profile, restrictions, and health goals
- Get AI-powered meal suggestions and explanations (using local LLM via LM Studio)
- Find nearby healthy restaurants and grocery stores (Google Places API)
- Track progress and celebrate achievements
Modern, user-friendly login page with social sign-in options and a welcoming introduction.
Personalized dashboard showing calories, meals, daily goals, streaks, and a summary of today's meals.
Track your weekly calories, protein, meals completed, and hydration with a detailed daily breakdown.
AI-generated daily meal plan with nutrition breakdown and a smart shopping list for easy meal prep.
Gamified achievements to motivate users, including streaks, protein goals, and hydration milestones.
- End-to-End AI Integration: Combines classic web dev with cutting-edge local LLMs for real, practical AI use.
- Privacy-First: All AI runs locally—no user data leaves your machine.
- Modular & Extensible: Clean, well-organized codebase makes it easy to add new features or swap components.
- Real-World Relevance: Tackles a meaningful problem—personalized health and nutrition.
- Modern Stack: Uses the latest in React, TypeScript, Flask, and AI APIs.
- Beautiful UI: Professional, responsive design with Material UI.
- Open Source Spirit: Built for learning, sharing, and community improvement.
- Explore Modular Code: Each feature (diet, nutrition, LLM, places) is in its own module for clarity.
- Add Your Own AI Models: Swap in any OpenAI-compatible LLM in LM Studio—experiment with prompts and outputs!
- Improve the UI/UX: The React frontend is easy to extend with new pages, themes, or components.
- API-First Design: All business logic is exposed via RESTful APIs—perfect for mobile or other frontends.
- Learn Best Practices: See examples of environment config, error handling, and scalable project structure.
- Contribute: Fork, branch, and PR! Suggestions, bugfixes, and new features are welcome.
- Great for Portfolios: Use this as a template or inspiration for your own AI-powered apps.
- User Profile Management: Age, gender, weight, height, activity, restrictions, health conditions
- Personalized Diet Generation: Smart meal plans tailored to user needs
- AI Meal Suggestions: Detailed, context-aware meal ideas from a local LLM
- Nutrition Analysis: Macro/micro breakdown for each plan
- Nearby Restaurants & Stores: Find healthy options near you (Google Places API)
- Progress Tracking & Achievements: Visualize and celebrate your health journey
- Modern UI/UX: Responsive, clean React interface with Material UI
- Frontend: React, TypeScript, Material UI, Axios
- Backend: Flask, Flask-CORS, Python, Pandas, Requests
- AI/LLM: LM Studio (OpenAI-compatible local server), Phi 3.1 Mini 128k Instruct (recommended)
- Data: CSV-based food database
- APIs: Google Places (for restaurants/stores)
graph TD;
User-->|Browser|Frontend[React UI]
Frontend-->|REST API|Backend[Flask API]
Backend-->|LLM API|LMStudio[LM Studio LLM]
Backend-->|Google Places API|Google[Google Places]
Backend-->|CSV|FoodDB[Food Database]
git clone https://github.com/yourusername/ai-diet-assistant.git
cd ai-diet-assistant# Create and activate a virtual environment
python -m venv venv
./venv/Scripts/activate # On Windows
# Install dependencies
pip install -r requirements.txt
# Add your .env file (see below)GOOGLE_PLACES_API_KEY=your_google_api_key_here
FLASK_ENV=development
FLASK_DEBUG=1
cd src/frontend/diet-assistant-ui
npm install
# (Optional) Add .env for API URL
# echo REACT_APP_API_URL=http://localhost:5000 > .env
npm start- Download and install LM Studio
- Download the model: Phi 3.1 Mini 128k Instruct (IQ3_M or Q4_K_M recommended)
- Start LM Studio and load the model
- Ensure the OpenAI-compatible API server is running at
http://127.0.0.1:1234
- Open http://localhost:3000 in your browser
- Sign up or log in (UI only, no real authentication)
- Enter your profile details and generate a personalized diet plan
- Explore meal suggestions, nutrition breakdown, and local healthy options
- Add more foods: Edit
data/food_database.csv - Change LLM prompts: Edit
src/llm_integration.py - Swap LLM models: Use any OpenAI-compatible model in LM Studio
- Add authentication: Integrate with Firebase/Auth0 or Flask-Login
- Deploy: Use Heroku, Vercel, or your preferred cloud provider
- LLM not responding?
- Ensure LM Studio is running and model is loaded
- Check that the model name in
src/llm_integration.pymatches LM Studio - Make sure the API URL is
http://127.0.0.1:1234/v1/chat/completions
- Google Places not working?
- Add a valid API key to your
.envfile - Check backend logs for errors
- Add a valid API key to your
- Frontend/backend connection issues?
- Ensure both servers are running
- Check CORS settings in Flask
- LM Studio for local LLM serving
- Phi 3.1 Mini for the model
- Material UI for UI components
- Google Places API
This project is licensed under the MIT License. See LICENSE for details.