PackMyFit: Your AI-Powered Personal Stylist and Packing Assistant

PackMyFit is your one-stop solution for managing your wardrobe, creating perfect packing lists, and discovering stunning outfit combinations for any occasion. This web application, written in Python using Flask and SQL Alchemy, leverages the power of Gemini APIs via Google AI Studio and the Open Weather API for weather-aware recommendations.

Features

• Digital Closet: Organize your clothing items effortlessly.
• AI Packing Lists: Generate customized packing lists based on your trip details and weather conditions.
• Itinerary Generation: Create detailed itineraries for your trips.
• AI Outfit Recommendations: Discover stylish outfit combinations tailored to your preferences and the weather.
• Weather Integration: Incorporates real-time weather data for optimal packing suggestions and outfit recommendations.

Installation

1. Clone the repository:

   git clone https://github.com/shreyasnbhat/packmyfit.git

2. Navigate to the project directory:

   cd packmyfit

3. Create a virtual environment:

   python3 -m venv .

4. Activate the virtual environment:

   source bin/activate

5. Install dependencies:

   pip3 install -r requirements.txt

Configuration

• API Keys: Manage your API keys in api_keys.py.
• Feature Flags: Control application behavior using the flags in flags.py.

API Key Setup

• Weather Data:

  • PackMyFit can obtain weather forecast data for your trips & events, but this will require an OpenWeather API key.
  • If an OpenWeather API Key is NOT available, dummy weather data will be used. Set WEATHER_TESTING=True in flags.py.
  • If you have an OpenWeather API Key, set OPEN_WEATHER_API_KEY in api_keys.py & set WEATHER_TESTING=False in flags.py to enable real-time weather data features.

• AI Features (Packing & Itinerary):

  • PackMyFit is an Gemini powered application and requires a Google AI Studio API Key for the AI Generated Checklists, Itineraries, Outfits etc.
  • If a Gemini API Key is NOT available, dummy packing and itinerary information will be displayed. Set LLM_TESTING=True in flags.py.
  • If you have a Gemini API Key set GEMINI_API_KEY in api_keys.py & set LLM_TESTING=False in flags.py.

Testing Data

The testdata directory includes sample data for testing the AI features:

• Item Images & Metadata: Over 15 items across different categories with stock images and descriptions stored in item_repository.csv.
• User Preferences: Default user preferences are defined in user_preferences.txt.

Directory Structure:

testdata/
    - user_1_item_1/   # Stores the testdata images for User ID: 1, Item ID: 1
    - user_1_item_2/   # Stores the testdata images for User ID: 1, Item ID: 2
    - user_1_item_3/
    - ....
    - item_repository.csv
    - user_preferences.txt

Re-Initializing the SQL Alchemy Database

To clear the database and re-initialize it on each server restart, set RE_INIT_DB=True in flags.py. Then, start the Flask server:

flask run

Logging In

A default user account is provided for testing:

• Username: testuser
• Password: abcd1234

Adding an Item to Digital Closet

15 dummy items accross different categories are pre-configured. You may add new items if you'd like that. Navigate to the Items tab and click on the plus icon on the bottom right corner to add a new item, add details about the items and photos if you'd like.

• Generating Item Metadata:
  • A Gemini API key is required for this feature.
  • PackMyFit can upload the images provided by you and analyze them to figure out care instructions, item color, material, brand name etc.
  • Click the "Analyze Images" button to perform this analysis.

Creating a Checklist & Itinerary

A dummy trip from San Jose to New York (August 11, 2024 - August 15, 2024) is pre-configured. You may also add/remove trips as needed.

• Generating Itinerary & Checklist:
  • A Gemini API key is required for this feature.
  • PackMyFit leverages your digital closet, packing preferences, weather data, trip description, date, time, itinerary to provide personalized and detailed checklists.
  • Click the "Generate Itinerary" button to create an itinerary.
  • Click the "Generate Checklist" button to create a packing checklist referencing your wardrobe items. Generated checklists contain sub-checklists for each bag such as Carry On Bag, Backpack etc. Each sub-checklist references the items you own and provides deep links to the digital closet.

Creating an Outfit

A sample Hiking event in San Jose (August 11, 2024) is pre-configured. You may also add/remove events as needed.

• AI-Powered Outfit Recommendations:
  • A Gemini API key is required for this feature.
  • PackMyFit leverages your digital closet, weather data, event description, time, and date to provide various outfit suggestions.
  • Click "Generate Outfits" to obtain outfit options. The application will also suggest items you might want to add to your digital closet to complete an outfit.

Contributing

Contributions are welcome! Please open an issue or submit a pull request.