Searching the body of a PDF plus having custom fields

When you use the api what you send is what gets stored, so you would have to send Swiftype the body text directly. If you use the crawler you could maybe make use of one of the meta data fields they do crawl. Can Swiftype index PDF documents?.