Churn Prediction Project
Developed a machine learning model with Python (Pandas, Scikit-learn) to predict telecom customer churn. Cleaned 7,000+ records, used Random Forest, achieved AUC 0.81, and identified key drivers with feature importance.
View Notebook View Feature Plot
SQL Website Customer Insights
Created an SQLite database of synthetic website visitor data. Ran advanced SQL queries to uncover conversion rates, average time on site by device, and campaign performance with detailed tables.
View SQL File View Table
Customer Experience KPI Dashboard
Built a Power BI dashboard to track support metrics: resolution time, % resolved, customer ratings, with DAX measures and interactive visuals by channel and agent. Highlighted opportunities to improve chat resolution times.
Download PBIX View README