Natera interview questions

There were three phone interviews: one HR interview, one non-technical conversation with a Data Scientist, and one technical phone interview with a Data Scientist. The technical questions mainly pertained to my background in probability, statistical inference, and numerical modeling. There were also problems to solve, in the vein of discrete probabilities, stochastic signal processing, means, covariances, etc. Overall, the phone interviewers seemed pretty nice. After the 3 phone interviews, was a day-long onsite, consisting of 5 or 6 technical interviews. Natera is an interesting company, and their data scientists come from a variety of backgrounds. Some are trained physicists, others electrical engineers, others mathematicians, and others computer scientists. Each one had his/her own brand of question, and some were quite unhappy when I couldn't solve "easy", "common", or "obvious" questions in the interviewers' respective fields. I got one question that dealt directly with Natera's toughest numerical modeling problems, but the domain was unfamiliar to me, and I gave a prudent first-pass answer, followed with "but I'd have to visualize the data to get a better sense." The types of technical questions I heard were just as varied as the interviewers' backgrounds, and it seemed like Natera does not really know what it wants in a Data Scientist. I did well with technical questions from the physicist, computer scientist, and electrical engineer, but poorly with questions from the mathematicians. The office seemed frenetic, disorganized, emblematic of a company experiencing growing pains. For a while, an interviewer and I walked around in circles looking for a spare room to conduct the interview, because they didn't have an electronic system to reserve conference rooms.