Knewton Data Scientist Interview Questions
& Process

I got referred to the Data Scientist role at Knewton by a former colleague. I left my previous Ed Tech job to get into Data Science and have 2 years of experience. I talked to the Knewton recruiter about the Data Science Team and their Data Scientist opening. Based on my background teaching Stats during my PhD as a Computational Biologist, I thought it was a good fit. I applied for the Data Scientist role. The process began with a Knewton Machine Learning challenge where you use student responses to questions to keep good ones in a test bank. It took about 10 hours plus some reading on Item Response Theory, and it was really fun and challenging. After I finished, the recruiter said the Data Team loved my submission and she'd set up a phone screen with a team member, emphasizing they wanted to move fast. The phone screen was with the Head of Data Warehouse and mostly involved questions about the company and role. While we talked, I started to suspect it was more of a Data Analyst role than Data Scientist. I asked about it but didn't get a clear answer. The technical part involved him asking me simple SQL queries verbally, very quickly, without any CoderPad, which was different from the scheduling email. After the interview, I emailed the recruiter to confirm if it was for the Data Scientist or Data Analyst role listed on their website. She called instead of emailing. On the call, she said: - "I'm sorry you were confused about the role you were applying for." (Despite all emails mentioning Data Scientist or Data Team, never Data Analyst). - "The Data Scientist role is for people with years of experience in Machine Learning." (Which made me wonder why they gave an ML challenge to someone not doing ML, and it felt insulting given the early stage). - When I questioned why I'd interview for a lesser role than my current one, she responded: "The Data Analyst role isn't a lower position than Data Scientist." I decided to withdraw from the process during that call. My advice if you're interviewing with them: be wary of being misled. Even if told it's a Data Scientist position, remember my experience – they might be inexperienced or intentionally bait-and-switching to hire a Data Scientist for an Analyst role. It felt unprofessional and dishonest, so I wouldn't recommend wasting your time.

The process started with a phone screen by HR, followed by a coding challenge they sent over. They estimated it would take 'a weekend,' but it seemed much more involved due to its complexity. The task involved creating a recommender system and then generating a varied, randomized output from it. The stated goal was about distributing test questions to students, which, from my background in education, didn't align with standard or advisable evaluation methods. I decided against undertaking this excessively long, complex, and seemingly pointless task.

Had a quick Skype chat to go over my background and what I was looking for, and then got a coding challenge via email that I had 3 hours to complete.