Most of the reviewers seem butt-hurt by the process, so I'll try to give a more objective review. I highly doubt they're using their data challenge to fish for "free ideas", but it would be wise of them to make it not so closely related to Via. I applied online and had a 1 hour (which actually ended up being about 1.5 hour) phone interview with one of the data scientists. Most of the time was focused on one of my past research projects on HMMs which he seemed genuinely interested in. Clearly a smart guy and asked me really good questions about my project and then general applied stats questions about some problems Via might face. Conversation went well and I received their data challenge. The challenge was non-trivial on a massive dataset. It's not easy working on these data challenges after a long days work. They claimed they were "impressed" so I moved on to a 2nd round phone interview. Maybe it was because I was located in Boston, but this round should be in-person or at least over a video conference. I was distracted with work at my office and did not do well in explaining my approach to the data challenge and report. I was not surprised when the lady I interviewed with told me they're moving in a different direction afterwards. My suggestions: If the candidate's data challenge is truly good and of interest to the team, then invite them for an in-person interview so they can properly prepare and communicate their thoughts. At least make it a video interview. If the data challenge is not of interest, no further rounds should be conducted and they should provide a reason to the candidate.