You’re offline. This is a read only version of the page.
119
Hi,
I’m experiencing issues with the AI Builder. I’ve created a prompt to extract text from PDF files, which may contain standard text, scanned text (images), or a combination of both. However, some PDFs fail to process, and when I upload them, I receive the error message: "Something went wrong."
Could you help me understand what might be causing this issue?
Thanks!
Thank you for taking the time to answer my post!
Yes, the same files consistently fail, even when I retry. Here’s the prompt I’m using:
Extract the exact phrase "private limited banking" from the scanned PDF document file. The match must be case-insensitive but must be character-perfect. Do not allow any misspellings, extra letters, missing letters, or other variations. If there is any deviation from the exact phrase, do not return a match.
Unfortunately, I can't share the documents due to sensitive data. However, after some error handling on my end, here are my key findings:
File size isn’t the issue – I’ve successfully tested much larger files.
Low-quality scans still work – The prompt functions correctly even on lower-resolution documents.
Problematic files are visible – I can preview the failing files in the prompt testing window, and they appear fine.
Key observation: When using the pre-built AI model "Extract all the text in photos and PDF documents (OCR)," the files that fail in my AI prompt work perfectly. This model successfully extracts all text from the document.
#1
WarrenBelz
146,522
Most Valuable Professional
#2
RandyHayes
76,287
Super User 2024 Season 1
#3
Pstork1
65,890
Most Valuable Professional
Share
Report
All responses (
Answers (
