Batch OCR in practice · 26 September 2026
From 1,500 scans to 15 organized documents
A large collection, uploaded in one batch. Inksight processed the pages through OCR, preserved their order, and kept each folder in its own document. The resulting text was then reviewed.
The job
The collection contained 1,500 page images across 15 folders, with 100 pages in each. The task was to extract the text while keeping the collection organized: one document per folder, with every page in reading order.
From upload to review
- One batch upload. All 15 folders were dropped into Inksight’s New document screen together.
- Progress saved along the way. Processing resumed from checkpoints when needed, without uploading the whole collection again.
- 15 organized documents. Each folder became a separate document, with its 100 pages in order and OCR text available for review.
The results
| Pages uploaded | 1,500 |
|---|---|
| Folders organized | 15 folders → 15 separate documents |
| Page order | 100 pages per document, in the original order |
| OCR processing | 1,500 successful page requests |
| File integrity | All 1,500 uploaded files matched their originals |
For examples of handwriting recognition quality, see our handwriting OCR benchmark.
How the results were checked
The resulting text was reviewed in Inksight. Processing records were also checked against the original collection to confirm the upload, document grouping and page order.
All 1,500 uploaded files matched their originals. All 15 folders became separate documents, and each document retained its 100 pages in the correct order. The processing records show 1,500 successful OCR page requests.