Vocal Slice uses whisper for the speech-to-text step, then performs its own logic on the transcription and audio, search, take matching, audio slicing, etc.