Initial phone screen followed by technical phone interview with hiring manager. I think Glassdoor get the prize for worst data scientist interview question ever. It starts out fine, the hiring manager goes through your machine learning projects from work and talks about the technical details of its implementation. Finally you get to the coding portion of the interview. And you are asked to convert strings to binary algorithmically. Now if you are a computer scientist/CS graduate you may remember this. However, many data scientists do not come from a CS backgrounds, and I am pretty sure the hiring manager knows this by looking at your resume and actually interviewing you. It's odd because of the dissonance between the actual real world work and experience that is necessary to do data science work and the absolute irrelevance of the question. And it is not even one of the more common irrelevant algorithms/data structures coding questions that are asked during data science interviews, e.g. merge sort. It is really strange. One can have the relevant data science skills and experience, but be cut down by such a question. At the very least if they are going to ask an algorithms question it would be something like regression, why an interviewer would choose something like string-to-binary over something that is relevant to the job is hard to understand. However if you frame this issue in the context of bias and elitism then you may have an explanation.