The interview process began with questions focused on foundational ML/AI concepts, assessing understanding of essential topics like confusion matrices, which help visualize a model's performance in terms of true/false positives and negatives. You were also asked about recall, a metric evaluating a model's ability to capture all relevant instances, critical in contexts where minimizing false negatives is essential. Further, they covered ridge and lasso regression techniques, discussing their roles in regularization, with ridge applying an L2 penalty to reduce coefficient magnitudes and lasso using an L1 penalty that can drive some coefficients to zero, aiding in feature selection and overfitting control. Asked if worked with vector db.