I applied to the San Francisco office. The process started with a screening phone call with a recruiter, then a take-home ML exercise which you have to return within 2 weeks, and finally a technical interview which has theory questions as well as some coding.

What sucks is that what you hear about the company’s hiring process is true. Those people really ghost you after they interview you. Unless you reach out to them multiple times, they wouldn’t get back to you.

It sucks for a company that has a good reputation to be too lazy and careless to send a simple offer/rejection message after you spend days preparing for their interview.

For the take-home exercise, they send you a problem where you have to build a model to solve it.

For the interview, they ask you specific deep learning questions, as well as about plotting and analyzing probability distributions.