We want to translate original audio into different languages and produce audios from transcripts. However, we would love to have the ability to correct the translated transcripts automatically, as they are not connected to the original audio anymore. This feature would greatly enhance our workflow efficiency.