Status: Experienced, Pursuing MS in CS
Position: SDE I
Location: Seattle, WA
Date: Feb 29, 2020
Interview Process:
- Amazon Online Assesment
- Onsite Interview ( 4 Rounds )
Online Technical Assessment challenge:
There were 2 problems as per the standard to be completed in 90 minutes.They were:
Rotten Oranges
Number of Islands
So the challenge I faced was mainly time constraint. And to be honest, I felt technical rounds on-site weren't that easy for me. Make sure you get the approach right, even if all the test cases don't pass.
On-Site Interview process:
There were totally 4 rounds of 1 hour each. The format was, 20-35 minutes of behavioral questions and the rest 20-25 minutes alloted for problem solving on whiteboard or laptop. The behavioral I could recollect are:
Tell me about a successful project and why do you call it a success?
How would you convince a team member with opposing views to agree to you?
Note: To prepare for for behavioral questions please refer to Dan Coritor Videos. One of the best free preparation materials available online for Amazon.
1st Round:
Technical Question:
(Toughest round for me!)
The problem stated was not very vague. I couldn't quite relate to any approach or any problem that I had seen before, so I struggled a bit. The question was:
Get the Bacon number - a BFS problem. Given a list of strings, and that list contains unique actor names. If an actor has worked with another, they will be linked.
Suppose Chris Evans is given as input, and Chris Evans worked with Mark Wahlberg and Mark Wahlberg worked with Kevin Bacon, then distance to Kevin Bacon is 2. So your Bacon number is 2. It is a graph problem (Six degrees of Bacon).
https://oracleofbacon.org/movielinks.php
I couldn't solve it entirely, as I figured my approach was wrong from the beginning.
2nd Round:
(Hiring Manager round - with the amount of drilling, I figured this could've been the bar raiser round)
Again, first 25 minutes behavioral questions, and that was the pattern in rest of the rounds as well.
Technical Question:
Design a Google Photos like App.
- The REST APIs required,
- the choice of database,
- difference between SQL & NoSQL databases,
- why not SQL instead of NoSQL (since I used SQL DB)
- tables and column design of the database
3rd Round:
2 Behavioral questions and then 25-30 minutes of technical question.
Technical Question:
Given a list of strings, (or words), return a list of compound words whose sub components are present in the list. If not found return null. Initial explanation of the problem by the interviewer was pretty vague and I had a tough time, digging deeper into the specifics. Maybe that was intentional.
Example: List of words {high, way, highway, super, market, supermarket, superman}
Output: [highway, supermarket]
Not "superman" because “man” is not present in the intial list. Only if all the sub components are present, do you have to return the compound word.
4th Round:
(Again, another manager)
He made me feel comfortable and at ease. He seemed like an approachable guy, but I had almost lost by then and had given up. Because in the technical rounds, either my approach was wrong intially and then instead of changing the approach, I continued. That was a mistake. This was the first on-site interview and although I was prepared, this experience was least of my expectations.
He asked a casual behavioral question and he was pretty chill. Then moved on to technical.
Technical Question:
Given an array of numbers - including negative numbers, duplicates etc.
Input is this array and a target number to find. It could return a list of lists that added up to the target number. It could contain two numbers or more, or just one number. It was sort of like a combo of two sum and three sum, I guess. I approached it with HashMap and explained my thought process. Few test cases didn't pass, and he didn't seem convinced. But on the downside, I could not edit and fix my code as there was no time.
It was a must needed experience to understand my standing in terms of preparation and LeetCode has played a major role in this. I would like to thank the entire community and its team here who is taking each member of this community a step closer to their right job. Good luck to you all!
The recruiter called me up in the evening and conveyed the result that I already had figured out! Lol!
I would appreciate your help in solving the questions above. Thank you!