Extremely difficult, too many steps, and way too long. I was interviewing with KoBold over a 3-month period. I have 10+ years of specialized experience in the geosciences, 15+ years of experience, and a PhD. I was rejected because I was believed to lack "geophysical understanding." The interview process consists of 7 steps: 1. HR interview (Interviewer A) 2. Presentation of prior work with Data Scientist + discussion and deep dive (Interviewer B) 3. Take home project (1-week). They want a solved problem on a geospatial dataset, with a python notebook, a source code file that runs, instructions on how to run the file, unit tests, instructions on how to run the unit tests, and a report of thoughts and findings 4. Discussion of take home with Data Scientist (Interviewer C) 5. Live coding interview for finding ore in geospatial datasets. You are interviewed by a geologist (Interviewer D) and another data scientist (Interviewer E). The geologist gives a 15-minute powerpoint presentation on geochemistry and formation of certain metals, then you are handed geospatial datasets and a coder pad session. This was the worst interview in my opinion. They tell you that you don't need a geological background to apply, then expect you to ingest 15 minutes of new information and use that on a heretofore unseen dataset while you're being grilled by Interviewer E. Horrible setup and you are destined to fail 6. Live coding interview with yet another Data Scientist (Interviewer F) on machine learning. There is a coding component but they're more interested in your philosophy and approach to solving ML problems, and being aware of specific challenges with geospatial datasets 7. Interview with the founder (Interviewer G). Vibe check That's a total of 7 different people, 3 different kinds of technical interviews, and 2 live coding interviews. I should have said no prior to the live coding... I almost did. It's hard to be good at engaging with people while doing deep focus coding work, while also learning new information. It's also bad that there is no interviewer continuity in the process; I had great rapport with two interviewers, and OK rapport with four more, but had one less than stellar experience with one person. I don't know how any candidate succeeds at this, and they're most certainly rejecting a ton of great talent with this ludicrous approach.