The process consisted of three separate interviews on subsequent days with three separate Software Developers and Team Leaders at Bloomberg.

It primarily consisted of questions designed to ascertain my programming knowledge and ability to code on the fly. Questions varied from the specific, such as:

• How would you code a least common multiple algorithm?
• How would you detect a loop in a linked list?

To the more general, such as given a very abstract, high-level diagram of the Bloomberg Terminal, and asked how you would set up and improve the communication between the Bloomberg and its clients.

Since I primarily applied with a PhD in statistics, a few questions on financial mathematics were included, such as:

• What is the Black-Scholes model?
• What are its weaknesses?
• What models would you suggest?
• What do you mean by "fat tails"?

However, since it was obvious rather quickly my statistical knowledge was far fresher and more up-to-date than my coding ability, which had been dormant for around five years, they didn't spend too much time on the statistics side of the questions.

The interview ended with a brief one-on-one with a person from HR that asked me some of the standard questions you see on "online interview aids." Like, "Why do you want to work at Bloomberg?" and "I see you've got extensive teaching experience; do you plan to go into academia?" and "What other areas are you applying to other than finance?"

The bulk of the interview process, thus, contained mainly knowledge-based questions unrelated to what you find in most interview help books and resources.

How would you detect a loop in a linked list?

This was a technical question. Finding the answer was difficult since I forgot how linked lists worked, but the interviewer didn't mind helping out with hints and guiding me along. They weren't, obviously, interested in the exact answer, but how my thinking works.

How would you find bugs in a very large program?

This is one of the few questions I wasn't able to answer to my satisfaction, at least, since I hadn't worked with large programs and was still limited to small, simple debugging techniques.