Initializing secure environment…
Initializing secure environment…
Find the personal data in a document and remove it in one pass. The detector looks for Aadhaar numbers, PAN numbers, IFSC codes, GST numbers, card numbers, phone numbers, email addresses, passport numbers and dates of birth, outlines every match with a page reference and the matched text, and lets you keep or drop each one before anything is written. Approved regions are then permanently destroyed, not covered. Nothing is uploaded — which is the point, since the data is exactly what you do not want to send anywhere.
To auto-redact personal data from a PDF free, add the file and it detects Aadhaar, PAN, card, phone, email and other identifiers, showing each match for review. Approve the ones to remove and the data is permanently destroyed. Nothing is uploaded.
Run the detector, find the 12-digit match, and check the page before approving — a 12-digit number elsewhere on the page is not necessarily an Aadhaar. Redact the whole line rather than the number alone, because a masked number next to a name and an address identifies the person just as effectively.
Auto-redact finds the account number and the IFSC. It will not find the transaction history, so decide what actually identifies the account holder and redact that too. For a statement going to a landlord or a school, removing the transactions as well as the identifiers is the point of the exercise.
A ten-digit run in a table, a date, or a reference number can look like a phone number. That is exactly why every match is listed with its value and page. Approving all detections without reading them is how a legitimate date gets redacted from a contract.
Detect, review each match, redact, then run a second scan over the output and search the result for the values you removed. Two independent checks, one before and one after, is the difference between a redaction you can defend and one you hope is right.
It is precise on well-formed identifiers — a 12-digit Aadhaar in the right format, a 10-digit PAN, a 16-digit card number — and conservative on free text. Every match is shown with its value and page before anything is removed, because a redaction tool that acts silently on a false positive destroys data you needed. Review the list.
No detector is exhaustive. It finds identifiers matching known formats, not a bank account number written as 'my account' or a nickname. That is why Privacy Scanner exists as a separate pass with broader patterns, and why the honest workflow is: detect, review, redact, then scan the output again.
Not until OCR has run, because detection matches characters. OCR the file first, then auto-redact the result. A scan that was recognised badly may have misread digits, and a misread digit will not match the format — check the OCR output before trusting the detection list.
Yes. The characters are deleted from the content stream, not covered by a drawing. Search the output for a redacted value and it will not be found, which is the test that proves it.
Indian identifiers in particular — Aadhaar, PAN, IFSC, GSTIN — plus card numbers, phone numbers, email addresses, passport numbers and dates of birth. Phone and email patterns are the most likely to produce false positives, which is why each match is reviewable.
Yes, as a regular expression or a literal string, for data specific to your organisation — an internal employee number, a matter reference, a customer code. Custom patterns run alongside the built-in ones and appear in the same review list.
No. Detection runs in your browser against a file that never leaves the device. This is the tool where a server-side implementation would be worst, because the payload is precisely the sensitive content.
Yes. Free, no account, no watermark, and no upload. Every detector and every redaction runs locally.
More security & privacy — all free, no upload.