American Credit Acceptance Data Scientist interview experience

I started by talking through my background and projects, then I was asked to walk through a problem on how an ACA-style lending decision could be built using machine learning—moving from a default-rate mindset into concrete modeling steps. Questions ranged from identifying useful features for a default model (credit history, personal details, loan characteristics, and similar inputs) to working through an end-to-end pipeline like cleaning and processing data, building and evaluating a model, and then thinking about monitoring after deployment.