One phone tech screen covered both SQL and Python questions. They do the best job I've seen of hosting this tech screen/interview. They use CoderPad and actually let you run the queries/functions. For SQL, they provide a nice schema for you to reference. For coding (Python), they provide basic methods and reference materials for you to review if needed.

I did very well on the SQL questions, which are listed below. A plus is talking through what you plan to do before you do it:

• For each store, show the % difference in sales between 2014 and 2015.
• % of all customers that have purchased at least 1 product.
• List all customers that live in a state, ordered by the number of unique products they bought.
• Find the earliest born and last born customers, by gender, who have bought at least 1 product.

I did not quite do as well on the Python stuff. I was able to answer all of the Python questions, but I received feedback that I should have talked through and thought through all the test cases they provide, so as to ensure you are working around all edge cases before starting to code. I got 3 questions, but forget the 3rd one. The two I remember were:

• Implement count_words(input_str) function which returns the number of words from the input string (hint: the edge cases they provide means you have to implement it manually).
• Implement count_substr(input_str, sub_str) function which returns the number of times the sub_str occurs in the input_str.

On-site includes 4 interviews. One lunch interview is focused on team fit and finding out if you are passionate about the role/company. You'll want to be able to answer questions like:

• What product would you want to work on, why? What would you do if you worked on that project?

The other 3 on-site interviews are "full stack." They focus on product sense, data modeling, SQL, data visualization, and coding (Python). You can make it into more of a conversation, and it's a good thing to ask clarifying questions. You will have to write everything on a whiteboard throughout the conversations, including data models, viz, SQL, and Python functions. They will ask you to solve a specific business problem by coming up with a Python function.

From a guy that didn't get the job, it was still a good experience. They do a good job of making it into a much more realistic scenario instead of just "tell me a time when..." type questions. Good luck to you!

Given a list of tuples of movie watched times, find how many unique minutes of the movie the viewer watched. For example: [(0,15),(10,25)]. The viewer watched 25 minutes of the movie.