Freelance data scientist · All case studies

Who leaves, who is worth keeping.

A California telecom needed a clear view of high-value customers and churn risk. I joined Q2 2022 account metrics to zip-code population, ran the statistical work, and shipped models the retention team could actually use.

Client: California telecommunications study. Built by Dilshad Raza.

The desk could see cancellations after they happened. It could not see which accounts were both likely to leave and worth a save offer — or how neighbourhood density changed the picture. Account files and census geography lived in different rooms.

I merged customer churn metrics with zip-code population, ran exploratory analysis with t-tests and regressions, and trained logistic regression and random-forest classifiers to flag save-worthy risk. The handoff was a scored table plus the features that actually moved the needle.

Hire Dilshad Raza for similar freelance data science, machine learning, and AI automation in the UK, United States, and Australia.