Riverside's AI transcriptions have strong accuracy, even for recordings with imperfect conditions like some echo and background noise. See sample tests in the video below.
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.