Document capture pilot checklist
================================
Source: https://yenra.com/web-based-document-capture-system/
Prepared October 4, 2026.

Pilot owner: __________ Document type: __________ Date: __________
Received items/pages: ________ Expected pages per item: ________
Master copy location and access: __________________________________________
Fields to extract and consequence of errors: ______________________________
Required verification and acceptance criteria: ____________________________
Destination system and acknowledgement: ___________________________________

For each test item record receipt ID, pages, legibility, orientation, extraction,
verified fields, metadata, reviewer, exceptions and repository receipt ID.
Include faint print, skew, missing pages, duplicates and relevant languages.
Test authorized and denied access, retry behavior and duplicate delivery.
Keep received material and corrections traceable according to policy.

Fictional example: 20 two-page requests; one lacks a page and is held.
19 complete requests x 2 = 38 pages; 19 x 4 fields = 76 expected values.
72 values initially match: 72/76 x 100 = 94.7368%, approximately 94.7%.
This is a field-match measure for this sample, not whole-document accuracy.
Four mismatches require correction; the incomplete request remains unresolved.

Exception ID: ______ Item: ______ Cause: ______ Owner: ______ Action: ______
Resolved and verified by/date: _____________________________________________
Do not equate a successful scan with authorization to destroy the source.
OCR reference: https://tesseract-ocr.github.io/tessdoc/ImproveQuality.html