Apple Software Engineer Interview Questions

Before designing it, I would clarify what data needs to sync, how quickly it must appear on other devices, and how conflict resolution should work when devices are offline. At a high level, I would keep a local state store on each device, sync changes through a backend service, and resolve conflicts using timestamps or domain-specific merge rules. I would also think about offline edits, retries, and background synchronization so the user experience stays smooth. The trade-off is between consistency and responsiveness, and I would usually prefer a design that feels instant on the device while still converging correctly across devices. I would also mention privacy and encryption because that is essential in an Apple product.

I would first think about splitting the sentence into words and reversing the list of words, which is the simplest solution. If I need to do it in place, I would reverse the whole string first and then reverse each individual word. The reason I choose this approach is that it avoids extra space if I have a mutable buffer and still runs in O(n) time. I would also mention edge cases like multiple spaces, leading and trailing spaces, and an empty string. If the interviewer wants a cleaner version, I would explain how I would normalize the spaces first.

I would compare the two trees recursively. If both nodes are null, they match; if only one is null, they do not; otherwise I compare the values and then recursively check the left and right children. The reason I choose recursion is that tree structure naturally fits recursive comparison. The time complexity is O(n) where n is the number of nodes compared, and the space complexity is O(h) for the recursion stack. I would also mention that BFS could work too if I want an iterative solution.

I would first think about storing the nodes from one list in a set and then scanning the other list. That works, but if I want a cleaner O(1)-space solution, I would use two pointers that switch to the other head when they reach the end. The reason I choose the two-pointer method is that it aligns the path lengths automatically, so the pointers meet at the intersection if one exists. The time complexity is O(n + m), and the space complexity is O(1). I would also mention edge cases like lists with no intersection and lists that intersect at the head.

Before designing it, I would clarify the threat model, token lifetime, delivery channel, and whether the reset should invalidate all existing sessions. At a high level, I would generate a one-time token, send it to the user through a verified channel like email, and store only the hashed token server-side with an expiration time. I would also rate limit requests so the flow cannot be abused for spam or account enumeration. The reason I choose this design is that password reset is a high-risk path and needs to balance usability with strong security. I would also mention audit logging and session revocation after a successful reset.