This interviewer isn't just looking for "correct" answers; they are looking for System Reliability and Edge Case Handling. They want to know: Can you prevent a system crash when 1,000 people buy 1 item? Do you understand how to scale Kafka consumers without losing data? Can you write clean, optimized code under pressure?