I was reached out to by 3 different Amazon recruiters for 3 different positions in one week in May. Said no to one. However, for the other two positions the hiring managers also reached out to me personally and talked to me and described the position and responsibilities to me. After I agreed to interview for both the positions, the recruiter scheduled phone screens for both the positions. Apparently in Amazon, you can interview for two teams at the same time. I was a little iffy because I had rejected a ML scientist L5 offer in 2018. But, they still were willing to consider me.

Interview Details:

Phone Screen Round 1: One simple coding question on finding the kth largest element in an array. DL breadth covering when to use wide vs deep networks; whether GD/SGD always decreases loss; one LP question

Phone Screen Round 2: ML Depth, mostly focusing on my PhD research and publications. We discussed a particular paper of mine in great depth. It was more of an open ended discussion than the interviewer asking me questions.

Both my phone screens went pretty well. The lead recruiter then asked me if I would still like to interview for both the teams or just one team and if I had a preference. If I were to interview for 2 teams, there would be one talk + 6 rounds instead of one talk+5 rounds. So, I chose to go ahead with one team instead.

Virtual Onsite (Bound by NDA):

Round 1: One hour research talk; got asked a couple of questions may be throughout

Round 2: Bar Raiser: LPs and ML Applied problem solving pertaining to Recommendation Systems and Error constrained classification

Round 3: ML Breadth, Mostly related to properties of loss functions; optimization in general; robustness of DL algorithms; 2 LP questions

Round 4: 30 mins of 3 LPs, A coding question on strings. Due to video conferencing issues, I couldn't go through the test cases. I could only finish writing the code which was largely correct after discussing my approach.

Round 5: Hiring Manager round. 25 mins of LPs followed by a deep dive of one of my previous internship projects.

Round 6: ML Depth, Was very much like my phone screen round 2. The interviewer particularly discussed one of my recent papers in great depth. Questions ranged from from clarification questions to what ifs?

The recruiter reached out to me the very next day itself informing me that they would like to make an offer.