When transcriptions are generated, they often appear as a massive wall of text, especially when there is only one speaker. It would be beneficial to have the transcriptions automatically break up into smaller paragraphs to improve readability and make it easier to follow the content.