With the latest update, we have added the voice interface. It is very convenient (although it makes us a bit lazy). Loading .pdf and .txt documents is very fast and it is possible to reload them individually or reset everything. This is still a demo that aims to show how to use POWER-KI with GPT and now also how to provide a voice interface. I believe that starting from this, developers with good will can try to quickly create very interesting things. To convert speech to text, it is necessary to obtain a Microsoft Cognitive Services KEY.
Free an Open source as usual on GitHub.
