DokMine - Extraction & Redaction

DokMine removes personal data from documents — and actually removes it. Upload a PDF, Office file, image or email, and DokMine finds the emails, phone numbers, names, URLs and card numbers inside, then strips them from the file itself. What you get back looks like the original, minus the sensitive parts.

What it does

Drop in a file and DokMine scans it for personal information: email addresses, phone numbers, URLs, person names, credit card numbers and IBANs. From there you choose — extract the findings as a list for review, or redact them in place.

Redaction happens at the file level. DokMine deletes the text from the document's content stream rather than drawing a black rectangle over it. Move the shape, turn off the highlight, copy and paste the page: there's nothing underneath to recover.

Key features

• Format-preserving output — .docx comes back as .docx, .xlsx as .xlsx, PDFs as PDFs. Styles, tables and layout survive. Nothing gets flattened into plain text.

• OCR for images and scans — Reads screenshots, scanned pages and photos, corrects page orientation automatically, and redacts by editing the underlying image data.

• Honest about what it can't read — Files where OCR isn't confident are flagged for manual review rather than waved through as clean.

• Validated detection — Card numbers are checked against the Luhn algorithm, IBANs against mod-97, and company names are filtered out of person-name results. Fewer false positives to wade through.

• Consistent tokens — Matches become typed placeholders like [EMAIL_1] and [PHONE_2]. The same value always maps to the same token, so the document still reads properly and relationships stay traceable.

• Batch processing — Upload a whole folder. Each file gets its own result showing which entity types were found and where.

• Privacy by design — Uploads are used only to produce your output. Download links expire, and files are deleted from the server after processing.

Where the data hides

DokMine also checks the places most tools skip: tracked changes, comments, hidden sheets, speaker notes, slide masters and embedded images.

Why it stands out

Some redaction tools paint over text and leave the original recoverable underneath. DokMine performs true redaction, and the file stays usable afterwards — same format, same structure, same functionality.

It's also built to be checked rather than trusted blindly: images it can't read confidently get flagged, not silently skipped. Process documents you're authorised to handle, and review the redacted output before it goes out. No automated system is perfect.