Sentiment Analyzer API

📝 Overview

The Sentiment Analyzer API intelligently processes text input via a path parameter, utilizing NLTK's SentimentIntensityAnalyzer to extract negative, positive, and neutral scores. Based on these values, it determines the dominant emotion and classifies the sentiment in real time.

Designed for efficiency and scalability, this API automatically tests and deploys any code changes to AWS Lambda using the Serverless Framework. It seamlessly balances load with AWS API Gateway, ensuring high availability and performance under varying traffic loads.

With its serverless architecture, the API scales down to zero cost when idle (for Lambda alone). When used with API Gateway and S3, operational costs can reduce by up to 98%, making it a lightweight, cost-effective, and highly scalable sentiment analysis solution.

🎯 Use Cases

• Review Categorization: Automatically classifies reviews as Positive, Negative, or Neutral.
• Automated Customer Support: Negative reviews trigger AWS SNS notifications to notify the QA team and send a custom apology email to the customer.
• Recommendation System: Uses positive reviews to suggest the best-rated products or services.
• Scalability & Serverless: Designed to work seamlessly with AWS Lambda, ensuring cost efficiency and automatic scaling.

📊 Scalability Optimization & Cost Efficiency

• Uses AWS API Gateway for auto-scaling and load balancing under high traffic.
• Cold Start Prevention: Removes unnecessary dependencies after testing to ensure fast startup times.
• The API scales down to nearly zero cost for up to 1 million requests/month (including AWS services like S3, EC2, DynamoDB).

⚙️ Deployment Pipeline (CI/CD)

The API is fully serverless and automatically deploys when updates are pushed to the repository.

🏗 GitHub Actions Workflow

🔹 Triggers:

• 📌 Push to main or dev branches
  🔹 Deployment Jobs:
• Sets up Python 3.9
• Installs dependencies
• Runs security testing using run_checks.sh
• Deploys the API to AWS Lambda using Serverless Framework
• Prunes old deployments, keeping only the last two versions

🔐 Security & Testing Pipeline

Each deployment runs a test script that performs:

• ✅ Unit Tests (pytest)
• 🔍 Linting (flake8)
• 🗂 Import Organization (isort)
• 📏 Static Type Checking (mypy)
• 🛡 Security Scans (bandit)

🚀 API Endpoints

📡 1. Analyze Sentiment

GET /analyze/<text>

Parameters:

Parameter	Type	Description
text	string	Input text for sentiment analysis

Example Response:

{
  "positive": 0.8,
  "neutral": 0.1,
  "negative": 0.1,
  "dominant_sentiment": "Positive"
}

Conclusion

The Sentiment Analyzer API is a highly scalable, serverless solution designed for real-time sentiment analysis. It is ideal for customer feedback management, recommendation systems, and automated customer support workflows.

🚀 Built with Flask | Serverless | AWS Lambda | GitHub Actions | Passion :)

Developer and Maintainer

Ayush Pandey

🔗 Connect on LinkedIn
📧 Email: [email protected]