Reduced min test score to 0.2 because of switching to the smaller test model

Renamed tests for convenience

Added interview duration to feedback

Empty message handling

Updated progress options and waiting time

Hide record button while AI is answering

Comments and docstrings improvement

Improved audion input waiting, refactored the rest

Audio processing refactoring

Improved comments and docstrings

Speed up parallel tests execution

An ugly solution for the transcription delay problem

Removed unnecessary message box

Improved .env examples

Upgraded libraries