There are some examples floating around which we are using but is there an ETA for a proper api for the speech services ?