Document redaction is assessed on samples for format, targets and output; it is outside the standard image and video API pricing.
Image and video redaction are available now; text and document redaction start with a sample assessment.
PDF, JSON, CSV and TXT first; Word, Excel and PowerPoint confirmed separately.
Multi-page PDFs processed as one document with results per page.
Output format, layout and readability confirmed on real samples; exact layout preservation is not guaranteed by default.
For scoping; the actual scope follows sample validation.
Four document types that come up in scoping
Party names, ID numbers, contact details, amounts.
Names, phone numbers and ID numbers in registration exports.
ID numbers and addresses in published notices, partially masked by rule.
Personal information in employee records and HR documents.
Text layer first, OCR for scans
Scans and image-only PDFs are read first, then sensitive content is located.
Body text, tables and images are told apart and located precisely.
Results and output format are checked before integration; the parts you will review manually are agreed.
Describe formats, samples and volume; we assess and quote a deliverable scope.