Every answer has to be phrased in terms of the customer. Beyond that, the ds questions were pretty standard. Preparing answers where everything is clearly linked to the amazon leadership principles helps a lot.
Interview questions [1]
Question 1
How did the last product you worked on help the customer?
Asked questions about research (problems, difficutlies, metrics used)
asked general ML questions (Linear regression, regularization, etc)
Asked questions regarding leadership principles (make sure you have stories)
Simple coding question at the end
Interview questions [1]
Question 1
Asked questions about research (problems, difficutlies, metrics used)
asked general ML questions (Linear regression, regularization, etc)
Asked questions regarding leadership principles (make sure you have stories)
Simple coding question at the end
Phone screen with coding, virtual onsite. Mix of design questions, behavioral and technical breakout of your choice of dl architecture. One interviewers microphone was so faint I struggled to hear them, another had such a heavy accent it was only 10 minutes in that I realized what some of his words were as I slowly acclimated to his heavy heavy accent.
Interview questions [1]
Question 1
First coding was function to produce weighted probability sampler given a list of weights, second was least common ancestor in binary tree