Voice-AI FAQ

How to use Voice-live tool effectively?

For Best Accuracy, Select Your input Language! While the Auto Detect feature is powerful, the transcription model achieves the highest accuracy and speed when you specify the spoken language beforehand. If you know what language will be spoken, selecting it from the dropdown is highly recommended.

How to do transcription/translation in an online meeting or generally from a system audio?

For security and privacy reasons, web browsers cannot directly “listen” to your computer’s speaker output. To transcribe a meeting (from Zoom, Teams, etc.) or any other audio playing on your computer, you need to use a Virtual Audio Cable. This free software creates a virtual “loopback” device that routes your speaker audio to a virtual microphone, which you can then select in this web app.

Setup Instructions: Windows: Download and install VB-CABLE. In your meeting app’s audio settings, set the Speaker to CABLE Input. In this browser window, select CABLE Output as your microphone. macOS: Download and install BlackHole. In your Mac’s “Audio MIDI Setup,” create a “Multi-Output Device” that includes both your normal speakers and BlackHole. Set this as your system’s default output. In your meeting app, use default output. In this browser window, select BlackHole as your microphone. Linux: You can use PulseAudio Loopback Module. Load it by running the command: pactl load-module module-loopback. Use a tool like pavucontrol to route the audio from your meeting application’s monitor to your browser’s input.

Data Privacy

Are my conversations or usage data used for AI training or similar purposes?

No, your conversations and data are not used to train any AI models.

Are my audio files and conversations stored on your servers at any stage?

No, user audio files or BBB conversation are not stored at any stage on our servers. Once your audio is sent and you receive the response, the result is available in your browser. The results of the transcription/translation are saved on our data mover node which will be erased after 30 days. This gives the users the opportunity to download their results whithin 30 days. BBB transcriptions, which are written in Etherpad and use local MySQL to save data also will be kept for 30 days before they are erased (future service).

What data does Voice AI keep when I use the service?

We do not keep any conversations or audio files on our servers. Transcrition/translation results are stored in our data mover node for 30 days. BBB transcriptions, which are stored in local MySQL also will be erased after 30 days (Future service). We also record some usage statistics to monitor the load on our service and improve the user experience. This includes usernames, timestamps, and the services that were requested.

Converting Larger Audio Files to FLAC Format

If your file is too large, you can use tools like FFmpeg or Audacity to convert it to FLAC, which is a lossless format. For example, using FFmpeg:

ffmpeg -i input.wav -vn -acodec flac output.flac

Availability

My institution is interested in using Voice AI. Can we advertise it to our users? Would you be able to handle an additional load for XXX users?

For large institutions, please contact us directly at info@kisski.de.

Are Voice AI services for free?

Voice AI services that are accessible to a user with an AcademicCloud account are for free.