Initial call with recruiter, a standard recruiter screen.

My initial recruiter was awesome, but the second one I was passed on to was not so much. I felt like I was very much a name on a screen at that point.

Technical phone screen: one programming question followed by a barrage of ML theory and probability questions.

Then a virtual on-site, which consisted of 5.5 rounds/hours, including:

• Coding
• Data coding
• Data product design
• ML theory
• HM

Overall, it was a pretty terrible experience. The prep material and the reality of the interview content did not align well.

My very first interview, the interviewer was 15 minutes late due to a technical mishap. Instead of rescheduling that portion, they forced us to go ahead, which sucked (a coding round with additional stress) and really set the tone for the rest of the interviews.

Overall, the interviews were of average difficulty but were quite different from expectations set by the prep material and also by industry standards. I'm talking about some really elementary questions, but I lacked prep in those areas because they weren't listed in the prep guide or aren't commonly asked in standard MLE interviews.

It felt like most interviewers wanted you to fail, which was also very demotivating and disappointing. There's also this weird notion between data mining and AI, which is honestly just confusing for candidates, and there should be more clarity provided.

One positive was the HM round; the HM was nice and we were able to have a good conversation.

Study probability. They won't tell you to, but do it.

In my experience, a 'data product' is not the same as an ML system design.