A recruiter reached out about a new Data Engineering team being set up at Jane Street.

The first stage was a call with two managers (one of them an ex-trader) who were setting up the Data Engineering team. It was a fairly straightforward stage: run down your CV, talk about your experience, projects, etc.

The second stage was a coding round, specifically pair programming, with one exception. You didn't have to execute your code. You still had to write real code (rather than pseudocode), but the focus was on evaluating your thinking. There was only one challenge to code up: implement a Connect N game with two differences:

1. It can be of arbitrary size – infinite rows and columns.
2. Tokens are added from the bottom rather than the top.

The challenge was broken down into stages. If you implemented a stage, they revealed the next one. The interviewer said that the challenge is infinite; even if you implement all the stages, they will make up something just for you. The limitations are the time (1 hour) plus the imagination of the interviewer.

• Stage 1: Implement the class with the board. Implement the move function.
• Stage 2: Check the win condition vertically only for N consecutive tokens.
• Stage 3: Check the win condition horizontally for N consecutive tokens. They don't expect you to finish this part but want to hear about how you would approach it and code it up.

What kind of data did you work with?

How is your Data Engineering team set up at your current place?

How do you measure business impact?

What do you think of Airflow?

If you were designing an orchestrator, what would it look like?

Implement a ConnectN game with two differences:

1. The board can be of an arbitrary size – infinite rows and columns.
2. Tokens are added from the bottom rather than the top.