I applied in-person. The process took 4 weeks. I interviewed at Amazon (Seattle, WA) in Jan 2018
Interview
The potential manager reached out to me first. After an informal call with him checking whether our interests match, I had 2 phone screens with researchers people on that team. In the first one, I was asked machine learning/deep learning theory, with a practical task to program an LSTM cell update; following was the coding round where I programmed a simple search algorithm for navigating in a maze.
In the second screening, we only discussed a specific machine learning problem (apparently the one interviewer was himself working on recently).
At the end, the tasks are of average difficulty (harder on the ML side, easier on the coding side), but the way they assess your work is rather unpredictable.
Interview questions [1]
Question 1
Why would you regularize a neural network? List all the ways you know how to do it
I applied online. I interviewed at Amazon in Feb 2020
Interview
This interview was for an Applied Scientist II - NLP position. It was a 1-hour phone interview including online coding and some questions for my own research, followed by questions on ML and DL.
Interview questions [4]
Question 1
[Online coding question] TwoSum. Return the indice of two numbers that sum up to a given target number.
I applied through a recruiter. I interviewed at Amazon (Seattle, WA) in Feb 2020
Interview
Two phone interviews, then on-site.
on-site: Presentation + 2 interviews + lunch + 3 interviews
The recruiters are professional and helpful. While, the team matching is a unequal information game between teams and interviewee.
Interview questions [1]
Question 1
a lot of behavior questions. also some very basic machine learning and algebra questions.