I applied online. I interviewed at Arity (Chicago, IL) in Jul 2018
Interview
Phone interview with two data scientists. They were late to the interview and did not leave me any time to ask them questions. I felt one already had it in his mind to go with another candidate, they were just going through motions.
Interview questions [1]
Question 1
Describe a data pipeline you think their company uses, from consumer mobile app to data science solution. Also, code the local max from a list of integers using python or other language.
I applied online. I interviewed at Arity in Mar 2024
Interview
After passing 4 interviews, the recruiter all of a sudden, stopped answering my emails and messages. The recruiter sent me for one final interview then moments later retracted the email. Then he stopped responding to my emails. Completely unprofessional!
Interview questions [1]
Question 1
Talk about your past experience using large scale data
The overall interview process is very thorough and somewhat difficult. The whole interview process took ~2.5 months (which is the sole reason I am marking this experience as negative).
The interview process starts off with a phone screening through HR that tries to gauge your interest and background experience.
The next step of the interview process is a quick programming/basic statistics/data science check. I got asked questions about how to calculate some basic statistical properties of data using a chosen coding language. I then had to explain a previous ML/data science/numerical analysis project. Finally they asked me about how I would work with data in a given situation. We then discussed the repercussions/different scenarios that could occur.
The final step in the interview process is a very long (I think 6 hours) interview process. This consisted of 2 standard behavioral interviews, a programming interview, and a data science interview. The behavioral interviews are once again pretty standard, and just try to grasp at your interest to best match you to a team.
The programming interview consists of many different programming questions to work through. The first question was a standard fizzbuzz like question, the next one was to write some code to print and describe some information from a distribution, and finally the last one I got to was to sort a list.
The data science interview consisted of how you would start by analyzing a situation. This involved talking about your reasoning for look for different data points, and different algorithms you could use. I was later asked to look at some charts/statistics about a given ML algorithm and analyze what is wrong with them and how it could be fixed.