Q: Audio + Speaker Recognition
Hey, very cool tool. 1) does the tool recognize the different speakers on a call? I'd think that's important so it can provide accurate insights on who said what. Ideally there would be a way to tag and name the speakers so it's clear which said what. Google Recorder does this very well. 2) Is the audio file saved along the transcription? If so, can it be downloaded for use elsewhere?
julian_hedy
Nov 23, 2024A: Hi Amirtyz,
Thanks for the questions! I'll do my best to answer them:
The real-time speech to text model that we use unfortunately does not support speaker diarization, but it's something we want to add as soon as the technology to do this on a phone is available to us. Google Recorder does indeed do a decent job with that, but Google has that tech deeply integrated into the Android OS and doesn't make it available to other apps.
That said, it only really matters if you want to read the transcript, since the AI does incredibly well at discerning who said what from the transcript in its analysis and summaries.
Currently we do not save any audio (which is by design). All audio is processed in real-time and immediately discarded for privacy reasons. That said, we have received several requests for people that want to save the audio so we do have that on our roadmap.