A CLI tool that automates the creation of voice datasets by searching for character videos, downloading them, and using speaker diarization to isolate specific voices. It leverages audio embeddings and cosine similarity to filter out background noise and other speakers, producing a consolidated audio file suitable for voice cloning.