About

This comprehensive machine learning pipeline analyzes 4,771 Zerve users and 409K events to predict success, engineering 6 advanced behavioral features (diversity score, engagement depth, session metrics, recency) that improve ROC-AUC to 0.9992, then segments users into five distinct personas (Power Builders, AI-Native Users, Active Explorers, Onboarding Users, One-Time Visitors) using K-Means clustering with silhouette score 0.688, producing a full-stack data science workflow with interactive visualizations, model evaluation charts, and actionable business insights.