CruzHacks2025

Team Members: Bodie, Luis, Paul, Judy Project Name: Spritz

SpritzAI – Affordable Alternatives to Luxury Fragrances

Spritz is a web application built to help users find affordable alternatives—or "dupes"—for high-end perfumes and colognes. Aimed at students, budget-conscious shoppers, and everyday consumers who love great scents without the luxury markup, the app uses AI and web scraping to deliver personalized recommendations. Users simply paste the URL of a luxury fragrance product into the app, and it returns similar-smelling, budget-friendly options along with images, detailed information, and database-generated comparisons.

---

How It Works

1. User Input: The user enters a URL of a high-end fragrance product.
2. Web Scraping: The backend uses Puppeteer to scrape the product page's HTML.
3. Fragrance Analysis: The HTML is sent to the Gemini API, which identifies the product name and classifies it into a predefined scent category (e.g., floral, woody, citrus).
4. Database Lookup: The backend queries a PostgreSQL database (running in a Docker container) to find exact or category-based dupes.
5. AI-Powered Results: If matches are found, they’re returned to the user along with product images, purchase links, and LLM-generated descriptions and comparisons.

---

Tech Stack

Frontend

• React.js
• React Router
• TailwindCSS (for styling)

Backend

• Node.js
• Express.js
• Puppeteer
• Gemini API
• PostgreSQL
• Docker
• OpenAPI 3.0

---

Application Structure

• client/: React frontend
• server/: Node/Express backend
• docker/: Docker Compose & PostgreSQL setup
• db/: Database schema and dupe data

---

🚀 Running the App

To run the app locally:

# Stop and remove existing volumes
docker-compose -f ./docker/docker-compose.yml down --volumes

# Start PostgreSQL DB
npm run db

# Start backend server and frontend
npm run dev

#Start frontend (make sure all required components are installed, look in json file)
npm start