About

This project analyzes user activity data from the Zerve hackathon dataset to identify patterns that drive long-term user success and retention.

I performed end-to-end analysis including:
- Data exploration and success definition
- Feature engineering based on user behavior (sessions, days active, event diversity)
- Machine learning modeling (Gradient Boosting, Random Forest, Logistic Regression)
- User segmentation and insights dashboard

The analysis identifies key predictors of success such as days active, session count, and event diversity, and provides actionable recommendations to improve onboarding, engagement, and retention.

This project demonstrates effective use of Zerve’s agentic analytics, modeling, and visualization capabilities.