Phone call with hiring manager to explain the role (15 mins) and phone call interview with individual on team (1 hour). Lots of behavioral questions and questions where you have to walk through the steps on how you would architect a data pipeline.
Interview questions [2]
Question 1
How would you determine if a user is asking about the population of a city or a state that have the same name?
I applied through an employee referral. The process took 2 months. I interviewed at Amazon (Santa Barbara, CA) in Jan 2020
Interview
Applied to Amazon via an employee referral. A week later I was given a phone interview and a week after that I was invited to an onsite interview. Overall the entire process was very quick and very prompt.
The phone interview was difficult and it included questions about how Alexa could potentially return answers for questions with multiple possible responses, what to identify in a dataset, and what my proudest and least proud moments were in my career. Additionally what my experience was with Alexa and do I use Alexa for Q&A.
The onsite interview was additionally difficult, but was a lot more interactive and fun than I was expecting. One of my interviewers was not enthuastic and seemed unhappy to be there, but otherwise everyone was a pleasure to be around.
One thing to keep in mind:
Amazon has a policy of not disclosing the compensation scale for a role. I reached out to my point of contact for a compensation scale and we had a phone call where we discussed what my base salary was and if it was within the ballpark for what was being offered.
Amazon claims that they base offer packages on your experience and what you bring to the team, so they do not offer a pay scale before the onsite interview.
Interview questions [8]
Question 1
Phone Interview:
If a person asks Alexa a question about the population of Washington, what could she do to know which Washington the user is asking about
Phone Interview:
Amazon is trying to understand if questions about cars are a good dataset to go after. If you're provided with a database of questions asked, what could you do to figure out how many are about cars
In person Interview:
Amazon is targeting residents in a particular country to see how can best serve them.
> Assuming you are allowed to research, what questions might you ask and what data might be important?
> Given the results of that research (provided by interviewer), what assumptions can you make about the residents in that particular country and how they spend their time and money
> Given an implementation time for different subjects, how would you roll out development of the new features
In person Interview:
Alexa is having trouble answering questions about a specific category.
> Given a list of questions which have not been answered, what could you do to find out which questions are about the specific category (queries, datasets you could source, etc.)
> What are some potential reasons that Alexa is unable to answer these questions
> Given a list of questions, find commonalities between questions which have all been unanswered or answered
In person Interview:
KNOW THE LEADERSHIP PRINCIPLES! Have at least 2-3 examples for each leadership principle. During my interview (5 people), all 5 people asked for an example of the same leadership princople. I don't think they did this on purpose, but I felt very silly with my one good example and other not so wonderful ones.
Think of different ways a leadership principle question may be asked, they 100% will come up.
I applied online. I interviewed at Amazon in Jul 2019
Interview
First round interview was over the phone. It began by asking a few behavioral questions, then immediately jumped into technical questions. They gave no indication they would ask technical questions right off the bat. They took a longer time than they said they would.
Interview questions [1]
Question 1
If customers were asking about cars and Amazon Alexa is not allowed to provide information on cars, how would be able to identify questions asked about cars from Alexa's database?