🧠 ReviewIQ — Amazon Review Intelligence Platform

ReviewIQ is a full-stack web application that analyzes Amazon product reviews and converts them into actionable insights using AI. It helps users quickly understand what customers love, hate, and expect improvements on — without manually reading hundreds of reviews.

---

🚀 Live Features

Vercel:https://review-analytics-beta.vercel.app/ render: https://review-analytics-j59q.onrender.com

🔍 1. Single Product Analysis

• Input: Amazon product URL

• Output:

  • ⭐ Average Rating

  • 📊 Sentiment Breakdown (Positive / Neutral / Negative)

  • 💡 AI-generated insights:

    • Why customers buy
    • Common complaints
    • Suggested improvements

---

📊 2. AI-Powered Insights

• Uses Groq API (LLMs like Mixtral / GPT-OSS)
• Converts raw reviews → structured JSON:

{
  "buyReasons": [],
  "complaints": [],
  "improvements": []
}

---

🎯 3. Sentiment Analysis

• Derived from ratings:

  • ⭐ 4–5 → Positive
  • ⭐ 3 → Neutral
  • ⭐ 1–2 → Negative

• Displayed using:

  • Doughnut chart (Chart.js)
  • Progress bars

---

🖥️ 4. Modern Dashboard UI

• Clean dark UI
• Product scorecard
• Insight cards
• Interactive charts

---

🏗️ Tech Stack

Frontend

• ⚛️ React (Vite + TypeScript)
• 📊 Chart.js
• 🎨 Custom CSS (dark theme)

Backend

• 🟢 Node.js + Express
• 🌐 REST API

AI

• 🤖 Groq SDK (LLM-based analysis)

Scraping

• 🕷️ Playwright / Cheerio / Apify (earlier iterations)

• Extracts:

  • Product title
  • Reviews
  • Ratings

---

⚙️ Project Architecture

review-analytics/
│
├── backend/
│   ├── server.js        # Express server
│   ├── scrape.js        # Review + product scraping
│   ├── ai.js            # Groq AI integration
│   └── .env             # API keys
│
├── frontend/
│   ├── src/
│   │   ├── App.tsx
│   │   ├── Dashboard.tsx
│   │   ├── components/
│   │   └── styles/
│   └── index.html

---

🔄 Workflow

1. User enters Amazon product URL

2. Backend:

   • Extracts ASIN
   • Scrapes product title + reviews
   • Computes sentiment
   • Sends reviews to AI

3. AI:

   • Generates structured insights

4. Frontend:

   • Displays analytics dashboard

---

📡 API Endpoints

POST /analyze

Request:

{
  "url": "https://www.amazon.in/dp/XXXX"
}

Response:

{
  "productName": "...",
  "avgRating": 4.5,
  "reviewsCount": 10,
  "sentiment": {
    "positive": 80,
    "neutral": 10,
    "negative": 10
  },
  "insights": {
    "buyReasons": [],
    "complaints": [],
    "improvements": []
  }
}

---

🛠️ Setup Instructions

1️⃣ Clone Repo

git clone https://github.com/yourusername/reviewiq.git
cd reviewiq

---

2️⃣ Backend Setup

cd backend
npm install

Create .env:

GROQ_API_KEY=your_api_key

Run server:

node server.js

---

3️⃣ Frontend Setup

cd frontend
npm install
npm run dev

---

⚠️ Challenges Faced

❌ API Limitations

• Apify free tier memory limits (8192MB)
• Paid actor restrictions
• Service downtime (junglee maintenance)

✅ Solutions

• Reduced API dependency
• Implemented fallback scraping
• Optimized API calls

---

💡 Key Learnings

• Real-world scraping ≠ stable APIs
• Need for fallback mechanisms
• Data pipelines require multiple sources
• AI works best with structured prompts

---

🔮 Future Improvements

• 🔥 Competitor comparison dashboard
• 📈 Revenue estimation
• 🧠 Keyword extraction (battery, camera, etc.)
• ⚡ Caching results (reduce API cost)
• 🌍 Deploy as SaaS

---

🧑‍💻 Author

Harish Vasamsetti

• Full Stack Developer
• Passionate about AI + Product Building

---

⭐ Why This Project Matters

Most users:

❌ Read hundreds of reviews manually

ReviewIQ:

✅ Converts reviews → insights instantly

---

🏁 Conclusion

ReviewIQ is a step toward building:

🧠 AI-powered decision tools for e-commerce

It demonstrates:

• Full-stack engineering
• AI integration
• Real-world problem solving

---

📬 Feedback

Feel free to contribute or suggest improvements 🚀