I’m using the Amazon Polly TTS system to generate MP3 files of each chapter of an audiobook. It occasionally mispronounces individual words. I then have to experiment with the spelling to get the right pronunciation. For example, “They bow to the audience.” has to be spelled “They bough to the audience.” because it pronounces ‘bow’ like the weapon. I usually re-record the entire sentence with the mispronunciation. What’s the easiest workflow in Audacity to swap out the sentences?
Replacing mispronounced words in a Text to Speech (TTS) MP3 output file.
See this mini-tutorial in the Audacity manual: