The article mentions OCR, but you're sending a PDF how is that OCR? Or is this is mistake? What if you send photos of the pages, that would be true OCR - does the performance and price remain the same?

If so this unlocks a massive workflow for us.