I applied through an employee referral for a data scientist position. A recruiter reached out to me and directed me to the Data Science Institute’s job postings; two weeks after applying an administrator reached out to me to schedule a phone interview with two data scientists from LLNL. I was given a list of dates and times that the interviewers were free. I scheduled a time accordingly, but on the day of the call at the time we agreed upon, one of the interviewers called me and asked me to reschedule the interview to the next week.
After going through another reschedule through the administrator, we settled on a new time. This time the interviewers called me and were in fact able to interview me. First however, I was asked to physically write down a dial-in number for a teleconference with the other interviewer. After dialing in, the two interviewers opened by asking me if I had any questions for them, bereft of any real introduction on their part. I asked them about their projects, their roles, how the dynamic is at LLNL with the computation team and the research groups, etc. Most questions earned a sentence or two in response. It did not seem as if they had any interest in being there, but rather had to answer out of polite obligation. I did not sense any desire to sell LLNL as a desirable working environment, and no desire to sell the work or the people.
Then, they turned to me and asked me questions about my resume, including academic machine learning projects as well as some work experience and skills-related things. This part of the call goes reasonably. About 45 minutes into the hour long call, one of the interviewers abruptly switches focus, and asks me if I have a computer in front of me. This is the first time he mentions needing one (before or during this interview), and the first time anyone in this process so far mentions that there will be a coding portion. This is fine, as I was expecting there would be one and did have a computer. He then asks me if I can write down, with a pen and paper, a URL to access a pair coding platform. Once there, I hear him ask himself “what was the question I wanted to ask him again?”. Then upon remembering he proceeds to ask me a “convert a base 10 integer to a string representation of base 7”. This question catches me off guard, and to be fair, I flub it, getting only a partially correct solution.
I am told at the end that “I’m on the right track”, but it’s not fully correct (I knew as much) and the answer they were expecting was recursive in nature. There is a halfhearted attempt to explain it, and the call is over. The call ends at 12PM and the administrator emails me a rejection the very next morning at 8AM. This was a disappointing experience because it had seemed that the interviewers were not interested or invested in this process, and decided that a “convert-to-base-7” question with a desired recursive solution is a relevant question to ask for a data scientist position. I understand that a justification is that the approach might just to be to see how critically one can think through this problem. It’s primarily a math problem though, with no relevance to data science, useful algorithms, or data structures, that I was given 13 minutes total, including writing down URLs, to do.
At no point did they even make a mention of the rescheduling inconvenience, and were clearly disorganized and unprepared. It was disappointing since from all other accounts LLNL employees (both people I’ve interacted with directly and those I haven’t) are reasonable, passionate, and pleasant people. It came down to two young interviewers that were disorganized and uninterested, seemingly learning how to interview potential candidates on the fly with seemingly little oversight. From what I hear, you get brownie points on your yearly review from your boss if you volunteer to actively partake in lab activities, such as interviewing. So, if you want to be a part of and to contribute to Lawrence Livermore’s expanding data science and machine learning movement to solve our nation’s complex and meaningful problems, definitely know how to recursively convert to base-7.
