About

This canvas performs end-to-end user success prediction for the Zerve platform using machine learning, analyzing 200k+ events across 4 pillars (depth, retention, complexity, upgrade intent) to segment users into 5 tiers (Dormant to Champion) and identify AI agent adoption as the strongest success signal. The workflow chains together feature engineering → composite scoring → ML classification (XGBoost with 0.9996 ROC-AUC) → comprehensive visualization and actionable insights, culminating in a deployable prediction API.