User Engagement-Based Institutional Churn Analysis in Educational SaaS: An End-to-End Business Analytics and Lakehouse Approach

This repository provides the processed anonymized Gold-layer dataset and implementation code supporting the research study "User Engagement-Based Institutional Churn Analysis in Educational SaaS: An End-to-End Business Analytics and Lakehouse Approach" The dataset contains institution-level behavioral operational metrics and binary churn labels derived from the AIO Class platform (CV Smart Edutek Solusi). The accompanying notebook demonstrates the Medallion architecture workflow, exploratory data analysis, and a class-weight-adjusted Random Forest classification.