Business Analytics System

An end-to-end Business Analytics platform that performs audio transcription, sentiment analysis, and analytics visualization, fully containerized using Docker and deployed on AWS EC2.

This system converts audio into text using Whisper, analyzes sentiment using a trained Machine Learning model, stores results in MySQL, and displays analytics through an interactive Streamlit dashboard.

---

Deployment Status

• Deployed on AWS EC2 using Docker containers
• Instance is stopped when not in use to avoid cloud charges
• Live demo available on request

---

Tech Stack

• Python
• FastAPI (Backend API)
• Streamlit (Frontend Dashboard)
• Scikit-learn (TF-IDF + Logistic Regression)
• OpenAI Whisper (Speech-to-Text)
• MySQL (Database)
• Docker & Docker Compose (Containerization)
• AWS EC2 (Cloud Deployment)

---

System Architecture

→ Audio Input
→ Whisper Transcription
→ Text Preprocessing
→ TF-IDF Vectorization
→ Logistic Regression Sentiment Prediction
→ MySQL Database Storage
→ Streamlit Analytics Dashboard

---

Features

• Audio transcription using Whisper
• Sentiment analysis using trained Machine Learning model
• REST API built with FastAPI
• Interactive analytics dashboard using Streamlit
• MySQL database integration
• Fully containerized using Docker
• Production deployment on AWS EC2
• Persistent database using Docker volumes

---

Project Structure

Business_Analytics_System/
│
├── api/        # FastAPI backend
├── models/     # ML model and vectorizer
├── database/   # Database related files
├── asr/        # Whisper transcription logic
├── nlp/        # Preprocessing logic
├── app.py      # Streamlit frontend
├── docker-compose.yml # Multi-container setup
├── Dockerfile  # Frontend container
├── requirements.txt
└── README.md

This modular design improves maintainability, scalability, and production readiness.

---

Docker Services

The system runs using 3 Docker containers:

• frontend → Streamlit dashboard
• api → FastAPI ML inference service
• mysql → MySQL database

Managed using docker-compose.

---

⚙️ How to Run

Build and start containers

docker compose up --build -d

Stop Containers

docker compose down

Access the Application

• Frontend Dashboard:

http://localhost:8501

• API Endpoint:

http://localhost:8000

---

AWS Deployment

This project is deployed on AWS EC2 using Docker.

Deployment steps:

• Build Docker images
• Push images to Docker Hub
• Pull images on EC2
• Run using docker compose
• Access via EC2 public IP Instance can be stopped when not in use to avoid charges.

---

Machine Learning Model

Model: Logistic Regression
Vectorizer: TF-IDF
Library: Scikit-learn

Files:

• models/vectorizer.pkl
• models/sentiment_model.pkl

---

Key Highlights

• Production-ready Docker deployment
• Real-time ML inference using FastAPI
• End-to-end pipeline from audio to analytics
• Cloud deployment experience with AWS EC2
• Database integration with persistent storage

Notes:- Actual datasets and audio recordings are not included due to privacy.

Author

Abhay Sharma
GitHub: https://github.com/abhaysharma-dev