Comment by Splines

Longer term the team should track down where those PDFs are coming from.

"Source --> PDF --> docx --> HTML --> json" is just nuts.