I applied through a recruiter. The process took 3 months. I interviewed at Meta in May 2021
Interview
I was reached out by recruiter via email and scheduled phone call interview. Next I had on-site interview where they gave enough time to prepare and prep-calls are the best. They give you a very brief mock interview process for product-sense.
Interview questions [1]
Question 1
Main focus should be on Product sense, SQL and python basics like List and Dictionaries.
SQL - You should be well aware of aggregations, how to do one table scan and provide aggregated values.
Product sense is something which will be covered in all your rounds.
Data Modeling - Focus how you design your Dimension and Fact tables. Make sure to have denormalized fact tables. Go through Data models of Uber, Doordash, LinkedIn (not like graph database) but at higher level.
Because SQL questions would be based on how you design.
Python - Basic questions like calculating average, traversing and sorting lists and dictionaries would be enough.
Recruiter reached out. Setup a 1 hr screening interview.
This was my 2nd attempt at this position. Although, I could solve all the 4 SQL and 4 Python questions (passed all the test cases) with little help from the interviewer except the last one where the interviewer had to give me the idea to solve it, got an email the next day saying I didn't make it to the next round.
As they don't give feedback after the interview, it is very difficult to judge what went wrong. My understanding is, even though you have the option to run the code and see the output, don't do it until you are 100% sure that your code will pass all the test cases. Although the interviewer said that you can run queries and see data, every run you make is counted against you.
I would suggest don't put your heart and soul into preparing for the interview. It's ultimately a human interacting with another human. Even if you perform well in the interview, a lot of other factors can affect your candidacy.
Interview questions [1]
Question 1
SQL Data schema was provided. SQL to be written that used left joins, nulls in filter, count func., case stmt.
Python
create a histogram from a given list
flatten a list: [1,2,[3,4. [5],[6,7,[8,[9]]]]]
remove duplicates from a list
count words in a sentence
I applied through a recruiter. The process took 3 weeks. I interviewed at Meta (New York, NY)
Interview
Did a phone interview (5 SQL / 5 Python). Questions are same as in Glassdoor but edge cases are difficult to pass. Need to have good strategy in allocating time. The interviewer and recruiter are extremely helpful and nice.
Interview questions [1]
Question 1
Python (1) Count "char" in the string. (2) Replace "None" in an array. Codes are simple but passing edge cases are difficult. SQL several % calculation questions