I went through two stages: a technical screening and a coding session.

Screening was fairly standard - basic questions about Python and RDBMS, along with some questions about CQRS and Event-driven architecture. Nothing too difficult, the level of questions was reasonable.

Coding session turned out to be more challenging. The task was to implement a Load Balancer in 45 minutes. The format was interesting: the interviewer gives requirements iteratively, moving to the next one after you complete the previous.

What was confusing: the interviewer emphasized several times that the solution should be "production ready," but provided no feedback during the process. For me, production ready code means not just the logic, but also tests, logging, and error handling. This created confusion about priorities, especially when new requirements necessitated refactoring already written code along with tests, and time started running out.

Impression: The format with iterative requirements is decent in itself, but the lack of intermediate feedback and unclear criteria for "production ready" under time constraints made the task more difficult. It would have been more helpful if the interviewer had given at least some signals about whether I was heading in the right direction.

Recommendations for candidates:

• Be prepared for rapid coding under changing requirements
• Clarify upfront what exactly is meant by "production ready" in the context of interview coding
• Practice balancing code quality with implementation speed

Implement a Load Balancer with iterative requirements in 45 minutes, ensuring production-ready code quality.