A contracts lawyer told me to stop using PDF OCR, she was right

Last week I was prepping 400 pages of discovery for a case and kept getting garbage text from scanned docs. My buddy who does contracts work said ditch the OCR tool I was using and run everything through the court's e-filing system's built-in text layer instead. I thought she was overcomplicating it, but after trying it on one 40-page exhibit, the search hits actually lined up with the page numbers. Took me 3 hours less to find the key clause I needed for the motion. Has anyone else found that free county tools beat the paid software for this?

2 comments

2 Comments

kim.dylan19d ago

Hang on, did anyone think about what happens when that text layer data gets converted into a permanent court record? Once you file something through the e-filing system, the OCR text might get baked into the official docket. You could be locking in all those little errors the system makes, and then opposing counsel can point at your own filing to show the text is wrong. That's a whole different kind of headache than just having a messy PDF on your desk. The paid software at least lets you check and fix the text before it goes anywhere official. Seems like you'd want to keep the raw, messy version separate from the clean version you submit.

caseywest20d ago

County e-filing systems are clunky and built for filing, not for real review work, so relying on them for discovery is just trading one headache for another. Paid OCR tools give you way more control over batch processing and error fixing, which matters when you're digging through thousands of pages, not just one exhibit.