Your question is Memory-Efficient Duplicate Removal. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
Given a massive dataset, how would you write a memory-efficient algorithm to identify and remove duplicate records for Zephyr AI?