We have some really cool new features in v4, which we are rolling out at the moment – One of these is “Audio Redaction” – This allows you to highlight text in our new “Transcript” window, and the system will “bleep” the original file.
This means that if you need to disclose a recording for legal or regulatory purposes, you can quickly delete parts of the file you don’t want to disclose. Even if the transcript is incorrect (and as I keep saying, voice recognition is not magic…), because we keep the exact position of every word in the database, it will still redact the correct portion of the audio. No more trying to fiddle with playbacks and selecting what you hope if the right part.
Below is a demo based on the UBS “Rogue Trader” last recorded conversation. The preview function in the front end gives you a good idea of what the final file will sound like, but when you come to export the file, the backend finalises the process and gets the audio redaction exactly right. You can optionally receive the written transcript with the redaction in the text.