Know which customers are about to leave — before they do

A classification model that flags customers at risk of churning, so retention efforts go to the accounts that actually need them.

⚠️ This is an independent portfolio project built on public data — not a paid client engagement. Results are shown as-is, including known limitations.

Key results

Use case

Customer retention / churn risk scoring

How it works

Customer account, usage, and billing features are fed into a classification model trained to separate customers likely to churn from those who will stay, so retention offers can be targeted instead of blasted to everyone.

Have a similar problem in your operation?

Talk to us about your project