A couple years ago we were hiring data scientists, and started with a chat with the hiring manager, and then at some point a technical evaluation. We attracted business grads and others for the position (in addition to cs folks), and a lot of them talked a good game but couldn't do basic data science stuff. So we ended up switching the process to have some kind of table stakes technical evaluation up front, and then do the interviews.
I don't think it's ideal, but the filter has to be somewhere, and companies want to optimize hiring to cut people as quickly as possible rather than do a bunch of interviews and drop them later.
This is a good point that I overlooked and definitely agree is also present. The same thing is happening with other types of interviews - I have seen companies hiring now where the candidate is asked to record video answers to prompted question, that from what I remember are evaluated by some kind of machine learning. They can open up the funnel without having to do anything (except forgo candidates that either have some self respect or are not desperate for work)
I've been asked at least 3 times to do a similar thing, and every time I've just refused, it's a few steps too far for me to even participate in, even though I am almost completely out of options at this point.
lmao software houses in Eastern EU jump straight into day-to-day stuff, 99% of the stuff was normal development, that 1% was just for lulz, to check whether you heard about stuff.
0 algo questions.