My question is about training sources:

I signed up for the trial, but could only find manual ways of copying and pasting articles into the knowledge base. And I wonder whether the sources are restricted to that, or whether I could upload PDF/Word documents, Google documents, or use URLs to train the knowledge base.

In other words:

1.) existing web URL articles on my site that could be scraped and made part of the knowledge base
2.) an XML sitemap so that my chatbot knows everything that has been published on my website.
3.) Integrations with MS-Word, Google Docs or PDF files
4.) WordPress or other CMS integration — where a certain category or post types would comprise the articles to scrape and include in the knowledge base.

Hope that make sense, thank you!