Standard "Remove Duplicates" removes all but the very first occurrence based on the original ordering.

If you want to keep something else, for example the most recent row (which is pretty typical), you must do some workaround.

Out of all workarounds I've seen, this one by Brent Jones looks pretty good. It's a multi-step process where you do a little trick multiplicating the columns temporarily, creating a new True/False column indicating which rows you want to keep, and then filtering only True.

Brent's article goes step by step including screenshots.