PREPARE A LESS IDENTIFYING COPY

CSV Redaction & Pseudonym Workbench

Remove identifying columns, replace values with a redaction token, mask all but the last four characters or create repeatable keyed pseudonyms. Review a local preview before exporting.

1. Table & selected columns

CSV textOpen UTF-8 CSV

DelimiterColumns to change (1-based numbers)

First row is a header (keep it unchanged) Preserve empty valuesTransformationRedaction token

250,000 characters · 500 KB UTF-8 imports · 10,000 rows · 200 columns · rectangular tables only · 60-second worker limit.

2. Inspect the transformed copy

Try the example, then choose which columns to transform.

Pseudonymization is not guaranteed anonymization. Repeated values, locations, dates and other unselected columns can still identify people. Masking short values may leave them unchanged.

Complete transformed CSV

How to prepare a sharing copy

Import or paste a rectangular CSV table. Choose the delimiter and confirm whether row one is a header.
Enter the 1-based numbers of columns containing identifiers. Choose removal, redaction, masking or keyed pseudonyms.
For pseudonyms, generate a random private key. Equal values in the same column get equal IDs; different columns use separate domains.
Review the entire output, including remaining columns. Download the copy and a report without the private key or source values.

Important details

Quoted commas and newlines are parsed; output quotes all cells. Strings are preserved without numeric conversion. Row order remains unchanged. Pseudonyms depend on the exact value, column position and private key; changing any of those changes the identifier. No reversible lookup table is generated.

Masking preserves the final four Unicode code points and leaks length. Redaction and removal may destroy useful information. Unselected values and formulas are unchanged; use CSV Formula Sanitizer separately before spreadsheet import. Review any sensitive data carefully before sharing.

Using CSV Redaction & Pseudonym Workbench

Create a separate CSV copy with selected identifiers removed, redacted, masked or replaced by keyed pseudonyms.

Choose the delimiter, header option and 1-based column numbers.
Pick a transformation; generate and privately retain a key for repeatable pseudonyms.
Review remaining identifiers, then export a CSV copy and a key-free summary report.

Example

The sample redacts names and emails while retaining city and order columns for review.

Questions & answers

Is the output guaranteed anonymous?

No. Pseudonyms, repeated patterns, locations and other remaining attributes may still identify people. Review the complete dataset before sharing.

How are repeatable IDs generated?

HMAC-SHA256 uses your private key, column position and exact cell value. Equal values in the same column and with the same key yield equal IDs. The key is not saved automatically or included in reports.

What does masking retain?

The final four Unicode code points and the original length remain visible. Values four characters or shorter are unchanged; this is a convenience mask, not strong de-identification.

What tables are supported?

Rectangular UTF-8 comma, semicolon or tab-delimited data, up to 250,000 characters, 10,000 rows and 200 columns. Jobs stop after 60 seconds.

Are formulas removed?

No. Unselected values remain unchanged, including formulas. Use CSV Formula Sanitizer separately and verify spreadsheet import behavior.