For a machine learning engineer the model is the easy part. The interview is about what happens when it meets production: a good evaluation and a bad launch, a tail that is ten times slower than the median, a model you cannot reproduce a month later.
These are the skills in our catalogue that this role draws on. They are not the whole interview.