W Buy

Your AI-Driven Guide for Products, Places, and More

Comment 2

W Buy: Your AI-Powered Food/Travel/Purchase Companion

Inspiration

Online reviews today are bloated with sponsored content, bot-generated feedback, and manipulated ratings. Whether it’s Amazon, Yelp, or Google Maps—real voices are drowned out by paid noise.

  • Sponsored products disguised as top-rated
  • Bot-generated reviews that skew ratings
  • Review farms that artificially boost mediocre items or places

That’s why people turn to Reddit, hunting for real, unfiltered opinions. But scrolling through endless threads takes time. W Buy brings those insights to you - fast, reliable, and all in one place.

What It Does

W Buy is an AI-powered tool that:

  • Scrapes Reddit and Google Maps for authentic user comments and information
  • Sends them through Perplexity AI to summarize the most relevant insights
  • Lets you search for products, food, or locations based on category and place
  • Returns real recommendations with quotes and ratings from real people

No more scrolling. No more bots. Just the truth.

Bonus: The browser extension version currently supports Amazon.com, showing you the most relevant Reddit post and comments for the product you’re viewing—right on the product page.

How We Built It

Frontend (Next.js)

  • Built with React, Tailwind CSS, and React Icons
  • User interface for entering search queries, selecting location and category
  • Uses next/router for client-side navigation

Backend (Flask)

  • Provides /api/analyze endpoint for search requests
  • Uses:
    • PRAW to scrape Reddit
    • Google Places API for location data
    • BeautifulSoup4 for HTML parsing
  • Sends content to Perplexity AI API
  • Formats and returns results as JSON

Tech Stack

  • Frontend: Next.js, React, Tailwind CSS
  • Backend: Flask, Python, PRAW, Requests, BeautifulSoup4
  • AI: Perplexity AI API
  • Chrome Extension: JavaScript, HTML, CSS

Challenges We Ran Into

  • Trying to figure out a way to return Relevant data from Reddit
  • Scraping dynamic data while avoiding bot detection
  • Cleaning noisy Reddit threads for meaningful summaries
  • Failed to run a Web Scraper to find Product data from Amazon
  • Tuning Perplexity prompts for consistent results
  • UI optimization for displaying long-form responses
  • Getting the extension to sync correctly with Amazon product pages

Accomplishments That We're Proud Of

  • Combined Reddit + Google data into a single, user-friendly experience
  • Built a live extension that works seamlessly with Amazon
  • Designed a clean UI focused on clarity, trust, and simplicity
  • Completed a full-stack AI product with NLP in the backend

What We Learned

  • Building a Web Scraper and a Browser Extension for the first time
  • Prompt engineering for AI summarization
  • Efficient scraping and data cleaning techniques
  • Managing asynchronous UI updates and loading states
  • Importance of user trust when dealing with reviews and recommendations

What's Next for W Buy

  • Create user accounts to FAVOURITE generated posts and comments
  • Expand extension support to other shopping platforms (e.g., Walmart, eBay)
  • Improve location detection and Google Maps results
  • Train a custom model for Reddit/product summarization
  • Add natural language search (e.g., “Best pizza under $15 near me”)
  • Let users save and share recommendations
  • Develop a mobile version of the app
  • Enable CRON-based auto-scraping for fresh data

Built With

Share this project:

Updates