Well, the flip side of it was that in benchmarking our equivalent codebases for Unicode segmentation, my code was significantly faster, to both of our surprises

Well, the flip side of it was that in benchmarking our equivalent codebases for Unicode segmentation, my code was significantly faster, to both of our surprises as it turned out.