Vooks Capstone Project

I was part of a A data science capstone project analyzing customer behavior on Vooks, a children’s animated storybook SaaS platform, to improve trial-to-paid conversion and reduce subscriber churn. This was during the final semester of my Masters’ in Applied Data Science at Carnegie Mellon University.

In a team of 3 master’s students, we predicted trial conversion and subscription churn with machine learning models (LASSO logistic regression, XGBoost, balanced random forest) and survival analysis techniques (time-varying cox proportional hazard model) for a storybook streaming platform. We presented actionable business insights based on stakeholder input, data-driven recommendations, and qualitative industry research and developed skills in project planning, stakeholder management, and real-world applications of predictive machine learning models. Our technical deliverables can be found on this repo.