Free · No signup · Runs in your browser

Anonymize PII in your CSV.

Redact, hash, mask or replace sensitive columns with realistic fakes — per column, with PII auto-detected. Because it's local-first, the data you're scrubbing never leaves your browser.

01 · How it works

Three steps, then done.

Sharing a dataset but it has emails, names and IDs? This detects likely PII and lets you pick a strategy per column — redact, deterministically hash, mask, or swap for realistic fakes — entirely in your browser.

i. drop

Drop your CSV

Columns are classified and likely PII (email, phone, name, ID) is pre-set to a sensible strategy.

ii. choose

Pick a strategy per column

Keep, redact, hash (consistent token), mask (partial), or fake (realistic replacement).

iii. export

Copy or download

Get the anonymized CSV. Hashing is deterministic, so joins on hashed keys still work.

02 · Why ours

The perfect local-first job.

Scrubbing PII is exactly when you don't want an upload — and this never uploads.

  • 01

    PII auto-detected

    Email, phone, name and ID columns are recognized and pre-set so you start safe.

  • 02

    Five strategies

    Keep, redact, hash, mask or fake — chosen per column to fit what you're sharing.

  • 03

    Consistent hashing

    The same value always hashes to the same token, so relationships and joins survive anonymization.

  • 04

    Never uploaded

    All transformation happens in your browser — the whole reason a privacy tool should be local-first.

"Hashed the customer IDs, masked the emails, shipped the sample — without it ever leaving my laptop."
— an analyst sharing data
03 · FAQ

anonymizer questions.

What do the strategies do?
Keep leaves a column as-is. Redact replaces values with [redacted]. Hash produces a consistent non-reversible token per value. Mask keeps a character or two and stars the rest (emails keep their domain). Fake swaps in realistic same-kind values.
It's a fast, deterministic, non-reversible token — ideal for anonymizing keys while preserving joins. It is not a cryptographic hash; don't use it where a salted SHA-256 is required.
Yes — identical inputs always hash to the same token, so you can still join two anonymized files on a hashed key.
Never. Anonymization runs entirely in your browser; the file is never sent anywhere. Verify in DevTools → Network.
Yes — drop a .xlsx and it's read locally, then anonymized the same way.