Depends on whether you are lateral or college grad. For lateral it starts with your projects and what did you do. What techniques, why, how did you tweak the algorithms (if at all), why this technique over others. Also a lot of questions on how did you collect data, what changes or treatments did you apply. Also, how did you validate your model, was it implemented, business impact, etc.
I believe they us an E3 pattern rolled into an A5 heuristic. If the ML decides you are a choice based on K, then they deep learn your interview and run you through a filter. Good luck!
Nah it's applied ML engineer so they just use KNN