The process consisted of 3–4 rounds covering coding, machine learning, and math. Coding questions focused on arrays, matrices, edge cases, and time complexity. There was also a practical exercise building a pipeline on a small dataset, which included cleaning the data, handling missing values, selecting features, fitting a model, and evaluating performance. Theory questions covered probability, linear algebra, regression, regularization, and overfitting. Follow-ups involved explaining assumptions, debugging code, discussing alternative approaches, and walking through previous projects.