The interview was a highly intensive, multi-hour process divided into several stages. It began with a 90-minute data challenge, requiring the candidate to define a target variable and build a model. Afterward, the candidate had to present their project in detail, facing questions about their choices, model, and results. This was followed by individual technical interviews, where each interviewer focused on different topics such as logistic regression, decision trees, random forests, and deep learning. The questions were based on the candidate’s resume and the data challenge. Practical aspects like model training, deployment, and A/B testing were also discussed. Finally, a group session required the candidate to share a professional achievement, answer technical and product-related questions, and explain the rationale behind their work.