How would you optimize the search performance of a deeply nested data structure representing global exam locations at Cfa Institute?
