Looking for an AppleScript that can be added to Hazel, the script will remove the original OCR layer and then reapply the OCR

I have quite a few large databases in my Devon thing Pro office and needs some special attention.
When I initially imported the files, the copies were of poor quality, and many of the pages were skewed. All I can say is Thank God got MPU because after working 12 to 16 hours a day trying to get a handle on the nightmare I so looked forward to my drive home so I could learn from David and Katie. It took only a short time for me to realized that the Mac would be the only way I could muscle through this mess (by myself).
As I said earlier when I first imported the scanned documents into DEVONthink Pro office I unaware that the quality of the PDF would affect the OCR layer. I also was ignorant of the fact that if a page of a scanned document is sideways or slightly skewed the OCR possibly would not work correctly. When the dust settled, I had well over 85,000 files in 9 databases.
To combat the issue I was pulling out folders on to my desktop and then using the space bar to preview the file quickly….but then there was the issue of OCR layer. I made this Marco I also used Karabiner Elements App mainly cause Brett Terpstra (:beers:) said it was cool.

Is there an AppleScript I can plug into Hazel that will first clear the OCR layer and then reapply the OCR Layer

Thanks

Not an automation, but perhaps a starting point: Rose talks about converting a PDF to an image and back to get rid of a wonky OCR layer in this post.