OpinionFlow 🗣️

Read full writeup @ Dev.to

Real‑time, cross‑store product‑review intelligence
Bright Data Real‑Time AI Agents Challenge (May 2025)

Shoppers drown in thousands of fragmented reviews. OpinionFlow turns that chaos into clarity: it scrapes fresh reviews from Amazon and Walmart in real‑time, distills crowd sentiment with AI, and surfaces instant, source‑cited insights—so anyone can decide with confidence.

Powered by Bright Data MCP, FastAPI, PGVector, and Retrieval‑Augmented Generation (RAG).

Features

🔍 Smart Product Discovery - Search across Amazon & Walmart simultaneously
📊 AI-Powered Analysis - Extract and analyze thousands of reviews with sentiment intelligence
💬 Interactive Chat - Ask questions about products and get source-cited answers
⚡ Real-time Processing - Fresh review data with instant insights
🔗 Cross-Store Comparison - Compare products from different retailers side-by-side

Tech Stack

Backend:

• FastAPI (Python)
• PGVector for vector storage
• Transformers & Sentence Transformers for AI
• Bright Data MCP for web scraping
• Google Generative AI

Frontend:

• React (Next.js)
• Tailwind CSS
• Lucide Icons

Quick Start

Prerequisites

• Python 3.11+
• Node.js 18+
• PostgreSQL with PGVector extension

Backend Setup

1. Clone and setup Python environment:

git clone <repository-url>
cd opinionflow
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
pip install -r requirements.txt

2. Environment Configuration: Create .env file in backend directory:

BRIGHT_DATA_API_KEY=
BRIGHT_DATA_SERP_ZONE=
BRIGHT_DATA_WEBUNLOCKER_ZONE=
GEMINI_API_KEY=
HUGGINGFACE_API_KEY=
PINECONE_API_KEY=
PINECONE_DISCOVERY_INDEX=
PINECONE_ENVIRONMENT=
PINECONE_REVIEWS_INDEX=

3. Start Backend:

cd backend
uvicorn main:app --reload --port 8000

Frontend Setup

1. Install dependencies:

cd react-frontend
npm install

2. Start development server:

npm run dev

3. Open browser: Navigate to http://localhost:3000

Usage

1. Search Products - Enter a product name to discover items across stores
2. Select & Compare - Choose products from different retailers to compare
3. Analyze Reviews - Let AI extract and analyze thousands of reviews
4. Ask Questions - Chat with the AI about product insights, comparisons, and reviews

Contributing

This project is part of the Bright Data Real-Time AI Agents Challenge. Contributions welcome!

License

MIT License