Be able to fix the proposed transcription:
  • typos
  • wrong speaker detection