It was not easy to get an interview with Unbabel: after an initial 30min call with an HR person, and submitting my solution to a technical challenge, I was contacted two weeks later with the purpose of scheduling interviews with some engineers. There was a lot of back and forth, including them cancelling at last minute on more than one occasion. This was a red flag, in retrospect. This whole scheduling stage took more than a month.
In the end, I had two technical interviews, each with two members of different Unbabel teams. The first interview was quite unpleasant. It made me wonder after if I had wandered into the wrong meeting by mistake, or if I had submitted my application to the wrong role, or if someone had told them I was applying to a different role than what I did. Whatever it was, something was definitely off: the questions did not really match the role description.
One of the interviewers in particular (which I believe is the hiring manager for this position) seemed *very* bored, unengaged, and was obviously not keen on my submission to the technical challenge. Made me wonder why they bothered to interview me if it was not up to their standards. His questions were not really questions, more “gotcha” like observations with no interest in follow-ups. Didn't make an effort to engage in conversation or listen.
At one point I tried to describe vectorization as a way to improve my solution and they didn't seem familiar with the concept, conflating it with threading. At another point I was asked about my current role responsibilities. I mentioned design and implementation of software features to perform nlp tasks. One of the interviewers thought I was talking about feature extraction from data and seemed not to be familiar with the concept of a software feature, which made the interaction confusing. The other interviewer chimed in with an unrelated question about my interest in research (?) so I couldn't clarify things. The whole conversation was awkward and disjointed.
There were other red flags as well: I asked how they dealt with market bugs, how they conducted retrospectives, etc. He just stated that they don't do that because there are no official product releases, just code being pushed, so the customers don't notice or report bugs (?!!). Strange answer, which made me think root cause analysis is not one of their processes. Also made me wonder to what degree, if at all, customer requests drive feature development.
In contrast, the second interview with a different pair of interviewers from another team, was fairly normal. At any rate, it was obvious I would not be a good fit and would have a lot of trouble communicating with some team members. I declined afterwards to proceed further with my application.
