Sentiment Analysis ML API

A production-ready sentiment analysis API built with FastAPI, scikit-learn, and Docker.

---

Overview

This API analyzes the sentiment of a given text (positive or negative) using a Machine Learning model.

The goal of this project is to demonstrate:

• Building REST APIs with FastAPI
• Integrating ML models into production systems
• Writing clean, maintainable project structures
• Testing and validation
• Containerization with Docker

---

Tech Stack

• Python
• FastAPI
• scikit-learn
• Docker
• Pytest

---

Project Structure

• ml-api-sentiment/

  • app/

    • main.py # API endpoints
    • model/ # ML model
    • schemas/ # Request validation
    • services/ # Business logic
    • utils/ # Logging utilities

  • tests/ # Unit tests

  • Dockerfile

  • requirements.txt

  • README.md

Running the Project

1. Local setup

pip install -r requirements.txt

2. Run the API

2.1. Console

uvicorn app.main:app --reload

2.2. With Docker

Build the image

docker build -t sentiment-api .

Run the container

docker run -p 8010:8010 sentiment-api

2.2. Deployment (Render)

This project is deployed using Render. You can find a live demo in: https://sentiment-api-0zg1.onrender.com/docs

3. Test the API locally

Open:

• http://localhost:8010/docs

API Endpoints

• Health Check

GET /health

Response:

{
  "status": "healthy"
}

• Predict Sentiment

POST /predict

Request:

{
  "text": "I love this product"
}

Response:

{
  "sentiment": "positive",
  "confidence": 0.87
}

Testing

Run tests with:

pytest

Features

• REST API built with FastAPI
• ML model integration
• Input validation with Pydantic
• Logging of requests and predictions
• Error handling
• Rate limiting
• API versioning
• Unit testing
• Dockerized for deployment
• Cloud deployment (Render)

Future improvements

• Replace dummy model with a real trained dataset
• Add authentication (API keys / JWT)
• Monitoring (Prometheus / Grafana)
• CI/CD pipeline

Author

Built as part of a Machine Learning / AI Engineering portfolio.

Motivation

This project demonstrates the ability to:

• Bring ML models into production
• Build scalable backend systems
• Combine ML + backend engineering