Whispy: Adapting STT Whisper Models to Real-Time Environments