Phone screen: they asked me a simple Python question, then a statistics question, and then a Black-Scholes model question (I passed).
The next round was the tech interview, covering Gen AI and statistics questions. They told me to have paper and pen ready, and said they'd also ask me Python coding, and possibly a stress testing framework question.
The interview started with Gen AI — they asked me about chunking, parent-child chunking. After that they asked how I'd approach a stress testing case study. I answered all of this so-so.
For the stress testing question, he asked me about WTI and Brent, and I went straight to a t-copula approach. Then he had me share my screen and write two functions.
The first was factorial of n — I used an iterative approach. He asked, "Is this recursive?" I said no, and that I could write a recursive version if he wanted, but he said it was fine. He had me talk through the time complexity of the iterative approach, and I said O(n).
The next coding question was to write squares of n integers — no issues there.
After that he asked me about the variance of a new distribution formed by combining two normal distributions, in two cases — independent and dependent — no issues there either.
Then there were two probability questions: the first was the three-doors problem — one door has a car behind it — and I had to give a strategy and the probability of winning the car. No problem.
The last one was the eight-balls problem — one ball is heavier than the rest, and you only get to use the balance scale twice — what strategy do you use to find the heavier ball? I answered this one fine too. He asked if I'd done these problems before, and I said no.
Afterward I got an invite for a 45-minute Zoom interview.
Discussion
Loading comments…