I applied online. The process took 5 weeks. I interviewed at Amazon (Sunnyvale, CA) in Apr 2022
Interview
The first Screening Interview was with a member of the team. It was several technical questions about computer vision and one leetcode medium problem in coderpad that was relevant to CV (think matrix manipulation). I mostly solved it with a few hints. Technical questions were broad and did not get very deep.
The Onsite Interview was five 1-hour rounds.
The offer was very high (30% higher than comparable positions at other companies), but the horror stories of long work weeks and the danger of PIP steered me toward a lower paying role that I think I'd be happier in.
Interview questions [1]
Question 1
- First round was a ~45-min presentation on a prior project.
- One design question with the hiring manager seeing how I would approach the problem they are working on. This was the most technically difficult interview
- One leetcode medium question with a software engineer.
- Other rounds were a mix of behavioral questions and technical questions about CV/ML.
- Each round had multiple behavioral questions (maybe half the time or more). I could tell they wanted me to be specific on my direct involvement and were gauging both how well I handled certain situations and how I could explain my core values.
Overall not a very difficult interview, as long as you have solid examples for the behavioral questions/leadership principles. The technical aspect was actually far easier than several other non-FAANG companies I interviewed with.
I applied through a recruiter. The process took 4 weeks. I interviewed at Amazon (Philadelphia, PA) in May 2021
Interview
1 recruiter, 2 phone screens, 1 onsite with 5 rounds with different people. The interview questions include machine learning, coding and behavior questions. Such as the basic knowledge about SVM, clustering, and deep learning.
I applied through an employee referral. I interviewed at Amazon (New York, NY) in Feb 2022
Interview
Applied through a LinkedIn connection. I refused to interview with them several times, so it took several months. The phone screen was very reasonable. They asked about basic ML stuff and then a simple question on binary trees. Final interview was not that difficult, but I didn't have a lot of experience with deep learning, so I got rejected.
Interview questions [1]
Question 1
Mostly deep learning stuff. One ML design interview. One data structure problem on coder pad. Most of the interviews are focused on their principles.