Meta Senior Software Engineer, Machine Learning interview questions

Typical Meta coding interview, with problems involving data structures and algorithms. I had to solve 2 problems in 40 minutes without compiling/running the code.

Systems Design: Design Facebook Messenger --> be prepared to support every single decision you make and explain terms thoroughly. The interviewer will question your decisions and will try to break you so be firm on why you selected that design ML Systems Design: although the requiter told me that they will ask me to design a recommendation system or design the News Feed for Facebook etc the reality was completely different leaving puzzled throughout the interview. Nevertheless, the interviewer asked me about my past projects and then he/she selects one that was interesting. Interviewer and interviewee spend almost 40' breaking down the task into smaller parts, understand the difficulty of the project, discuss what could have been done better / other solutions / other data preparation techniques/approaches etc Be prepared to write the mathematical formulae of everything you are talking about. Coding assessment: Leetcode is your friend. I was asked in total 7 coding questions; 6 came from Leetcode/Facebook section. However, do not expect to have the most frequently ones asked. You need to explain your solution as you go. You will also be asked if there are any edge cases you have forgotten and other ways to solve this (faster / less space complexity). They rarely ask for space complexity but time complexity is a must for them. Behavioral Interview: be prepared to answer questions about your difficult times. For example, tell me about a time you had a conflict, tell me about a stressful time and how did you cope with that / what did you learn, tell me about a constructive feedback, describe a project that start as X and changed in the way into Y -- why did that happen and what was the role you played. Be sincere, just that.