Customer Review Analytics Dashboard 📊

A real-time customer review analytics platform that uses AI-powered sentiment analysis to provide actionable insights from customer feedback.

🌟 Features

• Real-time Review Generation: Simulates customer reviews with realistic data
• Multi-Source Integration: Handles reviews from X (Twitter), Instagram, Web, and Email
• AI-Powered Sentiment Analysis: Multiple sentiment analysis engines:
  • Google Natural Language API
  • HuggingFace
  • Ollama (Local)
  • Fallback (Rule-based)
• Advanced Analytics:
  • Sentiment distribution
  • Source analysis
  • Product performance metrics
  • Trend identification
• Alert Management: Proactive handling of negative reviews
• Interactive Dashboard: Real-time updates and visualizations

Architecture

flowchart TD
    subgraph Frontend
        UI[React Components] --> RG[Review Generator]
        RG --> SA[Sentiment Analyzer]
        SA -->|Engine Selection| APIs
    end
    
    subgraph APIs
        GA[Google NL]
        HF[HuggingFace]
        OL[Ollama]
        FB[Fallback]
    end
    
    APIs -->|Normalized Scores| SA
    SA -->|Real-Time Updates| UI

Loading

Application Logic

graph TD
    A[Review Text] --> B{Analysis Engine}
    B -->|Primary| C[Google NL API]
    B -->|Secondary| D[HuggingFace]
    B -->|Local| E[Ollama]
    B -->|Fallback| F[Rule-Based]
    C --> G[Score: -1 to 1]
    D --> G
    E --> G
    F --> G
    G --> H{Categorization}
    H -->|≥ 0.2| I[Positive]
    H -->|≤ -0.2| J[Negative]
    H -->|Other| K[Neutral]

Loading

🚀 Getting Started

1. Clone the repository
2. Install dependencies:

   npm install

3. Set up environment variables in .env:

   VITE_GOOGLE_API_KEY=your_google_api_key
   VITE_HUGGINGFACE_API_KEY=your_huggingface_api_key
   VITE_OLLAMA_HOST=http://localhost:11434  # Optional, defaults to http://localhost:11434
   VITE_SENTIMENT_ENGINE=fallback  # Options: google, huggingface, ollama, fallback

4. Start the development server:

   npm run dev

🛠 Tech Stack

• Frontend: React 18 with TypeScript
• Styling: Tailwind CSS
• Charts: Recharts
• Icons: Lucide React
• Build Tool: Vite
• Testing: Vitest (ready to implement)
• Data Generation: Faker.js

📊 Dashboard Sections

1. Reviews Tab

   • Real-time review feed
   • Sentiment indicators
   • Source tracking
   • Star ratings

2. Analytics Tab

   • Sentiment distribution pie chart
   • Review source analysis
   • Product performance metrics
   • AI-driven insights

3. Alerts Tab

   • Negative review management
   • Action items tracking
   • Response workflow
   • Resolution tracking

🎯 Sentiment Analysis Logic

The application uses a multi-tiered approach to sentiment analysis:

1. Primary Analysis:

   • Uses one of four configurable engines:
     • Google Natural Language API (returns score -1 to 1)
     • HuggingFace (converts label probabilities to -1 to 1 score)
     • Ollama (prompted to return score between -1 to 1)
     • Fallback (rule-based word matching, returns -1 to 1)
   • Each engine is configured to normalize scores between -1 (most negative) and 1 (most positive)

2. Categorization:

   if (score >= 0.2) return 'positive'
   if (score <= -0.2) return 'negative'
   return 'neutral'

3. Fallback System:

   • Automatically switches to rule-based analysis if the selected engine fails
   • Uses predefined positive/negative word lists
   • Ensures continuous operation if API services are unavailable

🔄 Review Generation

The system generates realistic reviews using:

• Predefined templates
• Product categories
• Weighted sentiment distribution
• Realistic user data
• Varied review sources

📈 Performance Considerations

• Efficient state management
• Optimized rendering with React hooks
• Debounced API calls
• Responsive design for all screen sizes
• Error boundary implementation
• Fallback mechanisms for API failures

🔒 Security

• Environment variable protection
• API key security
• Rate limiting consideration
• XSS prevention
• CORS configuration

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📝 License

This project is licensed under the Apache License - see the LICENSE file for details.

🙏 Acknowledgments

• Icons by Lucide React
• Charts by Recharts
• Fake data generation by Faker.js
• Styling system by Tailwind CSS