<<<<<<< HEAD

PriceAI - AI-Powered Price Comparison

An intelligent price comparison platform that uses web scraping, vector search, and generative AI to find you the best deals across multiple e-commerce sites.

Features

• AI-Powered Search: Natural language product search across multiple retailers
• Smart Recommendations: Gemini AI generates personalized shopping recommendations
• Vector Search: FAISS-based semantic similarity for finding truly comparable products
• Multi-Source Comparison: Scrapes and compares prices from Amazon, eBay, and more
• Instant Savings: Shows estimated savings and quality ratings for alternatives

Tech Stack

• Frontend: Next.js 16, React 19, Tailwind CSS
• Backend: Next.js API Routes, Node.js
• AI: Google Gemini API for recommendations and embeddings
• Vector Search: FAISS for semantic product matching
• Web Scraping: Cheerio for HTML parsing (with planned integration of ScrapingBee/ZenRows for production)
• Deployment: Vercel

Getting Started

1. Clone and Install

```bash git clone cd priceai npm install ```

2. Set Environment Variables

Create a .env.local file with your API keys:

```env GEMINI_API_KEY=your_gemini_api_key ```

Get your Gemini API key from Google AI Studio

3. Run Locally

```bash npm run dev ```

Visit http://localhost:3000 to see the app.

Project Structure

``` app/ ├── api/ │ ├── search/ # Product search endpoint │ ├── vectorize/ # Vector store management │ ├── similar/ # Semantic search │ ├── recommend/ # AI recommendations │ └── insights/ # AI-generated insights ├── results/ # Results page └── page.tsx # Home page

components/ ├── product-card.tsx # Product display component └── recommendation-card.tsx

lib/ ├── scraper.ts # Web scraping utilities ├── vector-store.ts # FAISS vector database ├── gemini.ts # Gemini API integration └── types.ts # TypeScript types ```

API Endpoints

• POST /api/search - Search for products
• POST /api/vectorize - Add products to vector store
• POST /api/similar - Semantic search
• POST /api/recommend - Generate AI recommendations
• POST /api/insights - Get AI insights about products
• GET /api/products/index - View indexed vectors
• POST /api/products/clear - Clear vector store

Deployment

Deploy to Vercel

```bash npm run build vercel deploy ```

Configure Environment Variables on Vercel

1. Go to your Vercel project settings
2. Add GEMINI_API_KEY environment variable
3. Redeploy

Production Recommendations

For production, consider:

1. Use a proxy service for scraping:

   • ScrapingBee
   • ZenRows
   • Handles anti-bot measures and rate limiting

2. Use persistent vector storage:

   • Supabase for vector storage
   • Pinecone for managed vector DB
   • Self-hosted Qdrant

3. Add caching:

   • Use Redis for caching search results
   • Implement ISR (Incremental Static Regeneration)

4. Monitor and rate limit:

   • Add request rate limiting
   • Monitor API usage and costs

License

MIT

Support

For issues or questions, open an issue on GitHub.

PriceAI

Product recommendation system

3aa010f443030d2f9fc59d124b7481dedb5b15f7