DealScout (TEAM LEXA)

DealScout is a simple hackathon/demo project that finds and compares product deals across popular Indian e-commerce platforms. It provides a backend API (FastAPI) that supports image-based product search (embedding + similarity) and link-based scraping for Amazon, Flipkart, and Meesho. A Vite-powered React frontend consumes the API for an interactive demo.

Features:

• Image search: Upload a product image to find visually similar catalog items.
• Link search: Paste a product link (Amazon / Flipkart / Meesho) and fetch basic pricing/offers.
• Simple catalog: In-memory product catalog with precomputed image embeddings for demo purposes.
• Lightweight demo stack: FastAPI backend, torchvision embeddings (ResNet18), and a Vite + React frontend.

Quick Start (Windows)

Prerequisites:

• Python 3.10+ (or compatible)
• Node.js 16+ and npm

Backend (API)

1. Create and activate a virtual environment from the backend folder:

cd backend

PowerShell (recommended):

python -m venv venv; .\venv\Scripts\Activate.ps1

2. Install Python dependencies:

pip install -r requiremnts.txt

3. Run the FastAPI server (development):

uvicorn app:app --reload --host 0.0.0.0 --port 8000

The API root will be available at http://localhost:8000/ and interactive docs at http://localhost:8000/docs.

Frontend (Vite + React)

1. From the repository root open the frontend folder:

cd frontend

2. Install npm packages and start the dev server:

npm install npm run dev

By default Vite serves the frontend on http://localhost:5173/ (check terminal output). The frontend is configured to call the backend API; ensure the backend is running.

API Endpoints (overview)

• POST /search/image — form upload file (image) and optional session_id form field. Returns nearby catalog items.
• POST /search/link — JSON body { "url": "<product-url>", "session_id": "<id>" }. Returns scraped product info when supported.
• GET /history?session_id=... — returns recent queries for the session.

Project layout

• backend/ — FastAPI backend, catalog and scrapers.
• frontend/ — Vite + React demo app.

Notes and limitations

• This project is a demo/proof-of-concept. The scraping code is synchronous via Playwright helpers wrapped in a thread executor — not hardened for production.
• Product embeddings are precomputed from sample catalog.py image URLs at startup. In production, store embeddings in a proper vector DB or cache.
• CORS is permissive (allow_origins=["*"]) for demo; lock it down before deploying.

Contributing

• Feel free to open issues or submit PRs. For larger changes, please open an issue first describing the plan.

License

• This repository does not include a license file. Add LICENSE if you want to pick a specific OSS license.

Contact

• Project: DealScout (repo owner: tripcoded).

---

If you want, I can also:

• Add shields (build, python version) at the top of this README.
• Create a requirements.txt with pinned versions or a pyproject.toml for better dependency management.