Our transcription and other intelligence are 100% machine automated. The accuracy depends on the quality of the audio/video (which depends on background noise, internet speeds, how far the speaker is from microphone). The typical accuracy for keyword search is between 70 - 80%, and a higher for slide identification. Currently, we only determine unique speakers based on the dial in, so if e.g. 3 people are in the same room we consider them to be the person who dialed in.
Articles in this section
- How accurate is the transcription, speaker identification and other annotations by Voicefox?
- What exactly does Voicefox record and process in a meeting?
- What are the benefits of inviting Voicefox to a meeting?
- Which conference systems does Voicefox work with?
- Is it legal to record meetings with Voicefox? Should I inform participants?
- Who sees the smart playback or can make highlights during the meeting?