I applied online. The process took 3 months. I interviewed at Meta (Londres, Angleterre) in Oct 2020
Interview
First, there was a Phone-in with the FB recruiter. It was good and they scheduled an online coding one-to-one interview with Engineer. It was comparatively easy and in 2 days I got the message that the interview was successful and there will be a virtual onsite interview with the FB ML Team in london.
FB Recruiters are really helpful while preparing for the interviews. They are the best!
The Onsite interview had 5 rounds - Behavioural, 2 coding rounds, System Design, and ML Design. Coding rounds are average with problems similar to leetcode, and we need to prepare well in advance. System design and ML design are more focused on how we will approach the problem in real life scenario.
Interview questions [1]
Question 1
Coding questions were similar to leetcode and other competitive programming
One on one with a FB employee. Interview was conducted over video chat. Two leetscode questions. Fairly standard. Went through time and space complexity. Medium to Hard difficulty (per leetscode).
Interview questions [1]
Question 1
What is the time complexity of this solution? What is the space complexity?
I applied through other source. The process took 6 weeks. I interviewed at Meta in Dec 2020
Interview
I have more than 8 years of experience in ML & AI. I am really good at what I do and I have awards and accolades that need a list. Facebook approached me to work with them on ML projects which to me was really exciting. The issue is that I did not get a single ML question during the whole interview process. Instead I was asked questions about things that I learned years ago and do not really have time to go over them thoroughly anymore. I am a seasoned practitioner and not someone who just finished school without experience. The fact that Facebook had me go through a process of reviewing and testing data structure concepts that are not even relevant to ML along with me putting a lot of time and effort for the technical interview makes it of the worst if the worst interviewing experiences I have ever had. And I actually solved the problems they gave with the optimal solution and I showed nothing but grace and class during the whole process. The fact that Facebook rejected me after that is still ridiculous to me and really confusing. My suggestion is that if you want people who work in ML, please keep the interview in ML. If I was asked one question in the area I wouldn’t have been so negative. But all the people who met me from Facebook did not have any ML knowledge. It was very disappointing for me and I felt really insulted.
Interview questions [1]
Question 1
Graph search questions and greedy search questions