Fairly basic there... chunking audio (websocket) and send them for translation sequentially.