Please consider to make the bot use the open source Hive ASR-dictionary that I've built, to clean the transcripts after they've been generated by Whisper (https://github.com/mp-hive/Hive-ASR-Dictionary)
It will significantly improve the quality of the transcripts. I've manually built this library of Hive-related terms and usernames of the course of 9 months, and I'm continually adding new terms. More info in the readme within the repo.
Added bonus: You won't be referred to as "Cal" everywhere ;)
That is a great addition. Have to improve the quality of the data so that it has relevance and context.