Wine Quality Prediction - MLOps pipeline
his project demonstrates a production-grade Machine Learning Operations (MLOps) pipeline for predicting wine quality. It goes beyond simple model training by incorporating data versioning, automated workflows, containerization, and cloud-ready deployment strategies.
Key Highlights
Automated Pipeline: Integrated DVC (Data Version Control) to manage and automate the ML lifecycle (Ingestion -> Validation -> Transformation -> Training -> Evaluation).
Experiment Tracking: Leveraged MLflow to track hyperparameters, performance metrics (RMSE, MAE, R2), and model versioning.
Interactive UI: Custom-built Streamlit dashboard for real-time inference and one-click model retraining.
Production-Ready Deployment: Containerized the application using Docker and configured for AWS EC2 deployment.
CI/CD: Fully automated delivery pipeline using GitHub Actions.
Technology Stack
Languages: Python
Machine Learning: Scikit-Learn (ElasticNet Regression), Pandas, NumPy
MLOps: DVC (Data Version Control), MLflow (Experiment Tracking)
Web Framework: Streamlit
DevOps: Docker, GitHub Actions, AWS (EC2 & ECR)
Configuration: YAML based modular setup
<a href="https://github.com/aditya0589/wine-quality-mlops.git">View My GitHub Repo: https://github.com/aditya0589/wine-quality-mlops.git</a>
Key Highlights
Automated Pipeline: Integrated DVC (Data Version Control) to manage and automate the ML lifecycle (Ingestion -> Validation -> Transformation -> Training -> Evaluation).
Experiment Tracking: Leveraged MLflow to track hyperparameters, performance metrics (RMSE, MAE, R2), and model versioning.
Interactive UI: Custom-built Streamlit dashboard for real-time inference and one-click model retraining.
Production-Ready Deployment: Containerized the application using Docker and configured for AWS EC2 deployment.
CI/CD: Fully automated delivery pipeline using GitHub Actions.
Technology Stack
Languages: Python
Machine Learning: Scikit-Learn (ElasticNet Regression), Pandas, NumPy
MLOps: DVC (Data Version Control), MLflow (Experiment Tracking)
Web Framework: Streamlit
DevOps: Docker, GitHub Actions, AWS (EC2 & ECR)
Configuration: YAML based modular setup
<a href="https://github.com/aditya0589/wine-quality-mlops.git">View My GitHub Repo: https://github.com/aditya0589/wine-quality-mlops.git</a>