Capital One Machine Learning Engineer interview questions

First, I talked with the hiring manager about the team and role. Second, I did a data science take-home challenge. (Some may claim the process takes 4 hours, but I doubt that number is anywhere close to the empirical mean.) Then I did a full day of onsite interviews (5 I think). I got feedback saying I did well during the interview; however, the offer was for a role below what we discussed in advance. It doesn't really matter if this was an intentional bait and switch or some sort of internal mixup -- in either case the hiring manager overpromised and underdelivered. This showed that I could not rely on the hiring manager. Based on this example and two earlier specific instances, I lost confidence in him. I recommend this: listen closely and trust your instincts. If you are a motivated and driven software engineer, I would advice against going to work at Capital One. I know many people in my network that recommend against it. Most talented people I've known at C1 have since moved on to other places. Look at it this way: if you are going to work for a for-profit organization in the financial services industry, they should pay competitively. C1 does not. C1 will claim they differentiate in terms of work-life balance, but such a claim is exaggerated and certainly not universal. Be sure to compare such claims against other companies you are considering -- many big companies I know aren't really that different. There is one positive aspect about C1: use their interviews as practice. The interviewers I met were nice people. The interviews aren't particularly challenging relative to top Silicon Valley interviews, so it can be a useful experience if you have a day to spare.

Interviewed for ML engineer role at C4ML in Virginia. 1-hour technical phone interview, and then on-site. The onsite interview had 4 sessions (ML, behavioral, whiteboard coding, and case study). ML: Basic ML questions around my PhD research (applied ML). Behavioral: Describe a situation when you had to do X. Coding: Standard whiteboard session. Case study: Discuss/design encryption.