Taylor.Ai

Inspiration

The goal was to reimagine the online shopping experience by combining AI, creativity, and automation. Instead of manually searching for clothing items and prices across multiple websites, the idea was to build an intelligent crawler that could instantly gather data and generate complete outfits based on style and context.

What it does

This project is an AI-powered web crawler that scans the internet to find, generate, and price outfits. Users can provide a description, image, or reference style, and the system automatically finds visually similar clothing items, assembles them into full outfits, and provides purchase links with real-time pricing information.

How it was built

A custom web crawler collects structured data from e-commerce websites and integrates it with an AI outfit generator trained on image embeddings and fashion metadata. The backend manages data collection and filtering, while the frontend presents outfit combinations in a clean, consistent layout to ensure all generated images and items are uniformly displayed.

Challenges encountered

• Standardizing images from different websites with varying aspect ratios, lighting, and backgrounds
  
• Managing anti-scraping measures while maintaining ethical and compliant data collection
  
• Accurately matching clothing items based on visual similarity and contextual style
  
• Structuring generated photos so each outfit appears cohesive and properly aligned
  

Accomplishments

A functional crawler was developed that autonomously searches the internet, identifies fashion items, and generates curated outfits complete with pricing and purchase links. A clean and structured visual layout was also achieved, making outfits easy to compare and visually appealing.

Built With

• cloud
• css
• fastapi
• firebase
• html
• javascript
• python
• restful-apis