Related to this task:
Once text documents are something that can be uploaded to the library, we should be able to train WisdomAI on the content of the docs.