IdeaSift

OCR users struggle to get reliable results from varied documents

People building OCR and digitization workflows encounter different failure modes across scanned PDFs, unusual layouts, accented languages, and poor-quality source material. They may need to compare engines, prepare pages, tune settings, or clean up text, but the evidence points to a lack of an easy way to evaluate and improve results on their own documents. A focused workbench could help without claiming to solve OCR for every language or document type.

For OCR developers and document-digitization practitioners. Mentioned from Mar 2024 to Dec 2025 on Hacker News and Stack Exchange.

6 different people described this problem in 6 separate discussions.

Week of 2026-07-13: 0Week of 2026-07-20: 0Week of 2026-07-27: 0Week of 2026-08-03: 0Week of 2026-08-10: 0Week of 2026-08-17: 0Week of 2026-08-24: 0Week of 2026-08-31: 0Week of 2026-09-07: 0Week of 2026-09-14: 0Week of 2026-09-21: 0Week of 2026-09-28: 0
0 mentions in the last 12 weeks
Indie fit
5.0/10
Pain
5.3/10
Frequency
7.0/10
Willingness to pay
0.0/10
Momentum
5.0/10
Who pays
Professionals
Competition
Medium
Build difficulty
Medium

What people said

Quoted word for word. Follow a link to read the whole discussion.

  1. Application for pdf deskew Looking for a tool that can deskew pdf pages. The pages are scanned images of text. The text (or all of it I'm interested in) is in two columns. This is prepatory for OCR. All the OCR tools I have tried get confused by the columns and put chunks of text in the wrong places.
  2. There have been such a large number of OCR tools pop up over the past ~year; sorely in need for some benchmarks to compare them. Would love to see support for normal OCR tools like tesseract, EasyOCR, Microsoft Azure, etc
    cdrini on Hacker NewsNov 2025Asked for a tool
Build brief

See what to build and who will buy it

  • 1 product idea with the smallest useful version and pricing
  • 4 places to find your first customers
  • 4 more quotes from people who have this problem
  • Current workarounds, existing solutions and risks