Contacted online, if I recall correctly, for a Data Scientist role. First HR phone call was pretty standard and amiable. Next, hiring manager interviewed on a Python exercise (to assess my ease of use with Python data structures, likely) and a series of increasingly tougher SQL questions, which weren't a problem for me. Next, a take-home was sent...an interesting and relevant problem on whisper message categorization which I thoroughly enjoyed solving, laying out, and automating with Python. No attempt was made during this time to assuage my inquiries about how they're making money and setup for the future. I even expressed skepticism about the business, hoping that someone passionate would make a strong case and present the potential challenges that I could tackle in the job, but no enthusiasm to convince me came across. I think, instead, my skepticism might have come across as a hurdle they didn't need, instead of a good thing that they can feed off of. Nevertheless, it seems they liked my take home project because they got back to me to schedule yet another programming interview... this time with a Server Engineer, who asked about my resume again, then phrased a tougher programming problem rather poorly (which I later discovered was a "medium" level challenge on hackerrank on queues using two stacks, or something very similar); while I'm familiar with queues conceptually, I was unable to extract a clear problem definition from the interviewer. The server engineer, who was a somewhat inexperienced interviewer despite no doubt being very smart, made many assumptions, I think, about the extent of software engineering knowledge a Data Scientist should have, and I don't think I can fault that, except the charge to level-set and baseline the expectations before diving right in. But I get it, some places need machine learning engineers who are more software engineers than data scientists, so that's fair. Perhaps be upfront about the expectations there. My problem then is this: If the whole criterion for the position hinged on the solution of one specific problem that really dives into the depth of software engineering nous, then why not have this test up-front, with the problem description directly ported from hacker-rank so that the framing of the problem remains pure? Make the extent of software engineering chops you need obvious from the get go, and filter for that. Furthermore, make an attempt to show some passion during the interview process and maybe assuage some concerns about the viability of a young startup (especially one that's got some negative press recently) that potential employees might have. Perhaps it's a reflection of a certain callousness about the product, or perhaps it's a proclivity to hiring those who don't ask such ponderous questions. Nevertheless, while I'm glad the process fell through, and I still remain dubious and uninformed about the product's viability, it did sound like the nature of the ML problems they take on are rather interesting. That was the exciting part that they failed to communicate! Got an email the next day saying they're not moving forward. No reasons given. Even asked for any constructive feedback but no response.