The process took 2 weeks. I interviewed at Hyperscience (New York, NY) in Aug 2017
Interview
The hiring process started with a recruiter reaching out to me on LinkedIn. Had an initial phone call. I was sent a hackerrank test. The hackerrank had 2 questions to solve in an hour. I was scheduled for another call with one of the engineers. This was the interesting bit. He started out by asking me a few introductory questions about my background. Communication skills were horrible. There was definitely a communication gap and he wasn't able to explain accurately what the ask was. He kept doing that for the next 30 minutes and asked me to write code towards the end. I was able to successfully complete the code in 10 minutes but they passed on my profile.
I interview for my company too and usually do these typical exercises but the face of the company shouldn't be the one who can't send the message accurately and precisely. As an engineer I feel frustrated that I wasn't assessed correctly because of someone else's inability.
Two rounds phone interview followed by one virtual onsite. The process is very smooth and HR is very responsive. The onsite contains multiple one-on-one meeting including machine learning, algorithm, system design and product design.
Interview questions [1]
Question 1
Theoretical machine learning questions and white board algorithm questions.
I applied online. I interviewed at Hyperscience in Jul 2020
Interview
First contact with HR person, with a quick interview of about 30 minutes, asked about my experience in general and talked a little about the company. Then the technical interview conversation was scheduled with a machine learning engineer from the company, this conversation was about general aspects of ML, talk a little and then ask a specific question. All very polite and helpful.
I applied online. The process took 4 weeks. I interviewed at Hyperscience (New York, NY) in Nov 2019
Interview
Phone screen with HR, then phone interview and on-site interviews. The whole thing was well-organized and pretty quick. Interviews had a fairly standard format, with 1 fit interview, 2 CS interviews and 1 deep learning interview