Riverside's AI transcriptions have strong accuracy, even for recordings with imperfect conditions like some echo and background noise.
In English recordings, repeated words appear in the transcript as spoken, so you can find and remove the exact words you want when you edit using the transcript.
You may encounter incorrect words if part of the conversation is too quiet to be understood or if multiple speakers are talking over each other (crosstalk). Keep in mind that the transcriptions may use the most common spelling for names or similar-sounding terms: for example, "Erin" may be transcribed as Aaron. You can correct the text in the editor. You can also add names and terms to your Dictionary so future transcripts spell them correctly.