Problem Statement:

Businesses face the challenge of identifying customers at high risk of churning (leaving) to proactively implement retention strategies and mitigate revenue loss

Objective:

Create a production-grade machine learning pipeline capable of predicting customer churn with high recall, enabling actionable business decisions and improvements in ROI

Process Pipeline:

• Data Ingestion from Azure Blob Storage/SQL
• Data Preparation using Azure Databricks/Pandas
• Model Training with Gradient Boosting Classifier
• Deployment via Azure ML and Azure Functions
• Drift Monitoring and Retraining Pipelines

Technologies & Models:

Gradient Boosting Classifier, Azure ML, Azure Databricks, Scikit-learn

Key Outcomes:

High Recall Rate, Production-Grade Pipeline, ROI Improvement