Arpita Rathi

At YASH Technologies, I analyzed customer data to identify churn patterns and retention drivers, then developed a churn prediction workflow that incorporated business costs. I also authored a 12-section Proof-of-Concept document with the analysis, methodology, findings, and recommended business actions.

On my Olist E-Commerce Analytics & Retention Pipeline project, I analyzed $13.22M in merchandise revenue across 96K+ delivered orders. I used SQL to segment customers, identified one-time purchasing and high-value customer retention patterns, and built a four-page executive Power BI dashboard.

For my Maven Telecom churn project, I modeled the cost of false negatives and false positives and selected a decision threshold to minimize modeled retention cost. I compared Logistic Regression with Random Forest and deployed real-time churn scoring in an interactive Streamlit app.