ML model deployment using docker, kubernetes; API deployment with FastAPI; and MLOps using MLFlow for water potability dataset

A hands-on MLflow project demonstrating experiment tracking, model training, and lifecycle management using Scikit-learn, XGBoost, and Dagshub integration.

Walk through getting the baseline model up to a proper implementation while gradually increasing the number of tracked objects.

Les entreprises perdent chaque année des clients sans toujours comprendre pourquoi. Ce projet vous permettra de suivre et gérer le churn grâce à MLflow, en versionnant les modèles et visualisant les métriques pour améliorer la fidélisation et le revenu client.

Build end-to-end DL pipeline for computer vision (Image classification) for “Chest Disease Classification from Chest CT Scan Images” and deploy Flask web app to AWS EC2 with Docker and CI/CD tool: Jenkins

Project looks to create a stand-alone MLflow model registry which sits on its own Azure Container Registry, using an image, connected to a blob storage (artifact store) and internal sqlite db (registry store).

A sample container of Mlflow/Python for deployment to AWS Lambda for ML model serving.

CentraleSupélec/OpenClassrooms Data Scientist 2024-2025 - Projet 7 (EDA & modélisation seules)

End-to-end machine learning pipeline for customer churn prediction on imbalanced data. Includes ADASYN oversampling, DVC for version control, and MLflow for experiment tracking and model management.