After audio has been transcribed, it would be very helpful to have a global find/replace to deal with words which aren't in the dictionary and things that are consistently transcribed incorrectly due to accents, audio quality, etc.