Capital One Principal Data Scientist interview questions

As everyone else mentioned, there is a background check (easy), hackertest (easy), technical phone screen (on machine learning), NYC taxi challenge (easy but time consuming), and then onsite. People are extremely friendly so I left with a positive experience. 5 interviews on site plus lunch. Their role play on-site is ridiculous though. They gave me 10 pages to look at and I had only 15 minutes! seriously? I need 1 hour to interpret 10 pages of data! There were two case studies. One about machine learning, and one like a youtube video they have posted. Overall, I enjoyed the company a lot. People were extremely friendly. Sometimes, it makes me think, how can someone score well in all these rounds of interview given the stress and brain freezing. I guess there are people who can do that!

A recruiter reached out to me and was very encouraging and excited. The role expected a publication history and capacity to conduct research. The position was nominally up my alley, so I enthusiastically pursued the opportunity. The process was long, but each step was easy. Several stages must be passed before even discussing the real role, which makes the process prone to sinking applicants' time with no payoff. The first step was an easy coding challenge to test coding efficacy. I then had a technical phone interview that occurred about 10 days later, which tested general data science techniques. I then spoke to the hiring manager two weeks later for an hour about some details about the role, which did not sound as expected. A data challenge was then issued on a generic Kaggle-esque task. I came in for an in person interview 5 weeks later, comprising more standard data science questions (basically word problems with linear equations and some faux-consulting role play involving decision trees and ROC curves). I did not receive feedback for two weeks following the interview. At this point, the only part of the interview that touched on the skills involved in my role was the casual call with the hiring manage. After a three month process involving no test of my capability to perform the tasks involved in the role, I was rejected for not having the experience necessary for the position. Without disclosing identifiable information, despite being in the field a relatively short time, I have published in top machine learning venues and developed, by a wide margin, a state of the art technique within the field for which I had applied. The idea that my experience had been determined in an interview process that involved no technical assessment of the capabilities for the position at hand is the definition of insanity. I would like to praise the positivity provided by the recruiter throughout the process; that said, Capital One desperately needs to shorten its process, ask relevant questions, and get their heads out of the sand if they actually want to attract talent. Don't waste applicants' time like this, and don't vet using irrelevant metrics and capricious intuition.