Currently, utterances can only be uploaded one by one manually. It would be nice if we can upload a csv file of all utterances.