Upload a file with real names, emails, or IDs. Choose how each column is anonymized — redact, hash, or replace with realistic fake data. Download a version that's safe to share with anyone. Nothing ever leaves your browser.
100% client-side. No data is uploaded. Your file never leaves your device.
Contacts, deal values, internal IDs — the data you need to hand off is the data that's most sensitive. Every column needs a different approach.
Four steps from sensitive export to safe file. Everything runs in your browser — no upload, no server, no waiting.
Drag and drop or browse for a CSV file. The file is parsed locally in your browser. Nothing is transmitted. You immediately see all columns with sample values so you know exactly what you're working with.
Each column gets its own treatment: keep it as-is, redact it entirely, hash it for join-safe anonymization, or replace values with realistic fake data (names, emails, phone numbers, company names, addresses). A live preview cell shows what the output will look like before you commit.
Toggle between the original and anonymized view of the first rows. Verify every column looks right. If a strategy isn't working the way you expected, go back and adjust it — no re-upload needed.
The anonymized CSV is generated entirely in your browser. Download it and send it to the agency, contractor, or teammate who needed the data — without worrying about what's in it.
Different columns call for different treatment. Pick the right approach per column — mix and match in the same file.
Leave the column unchanged. For non-sensitive columns like country, status, or product type.
Replace every value with [REDACTED]. Use when the column's presence matters but its content doesn't.
SHA-256 of each value. Deterministic — the same input always produces the same hash — so you can still JOIN two anonymized files.
Replace with realistic generated data. Choose the sub-type: name, email, phone, company, address, date, or number. The output looks real without being real.
Keep only year and month. Temporal trends stay intact, exact dates do not.
Round to the nearest 100, 1,000, or 10,000. Revenue ranges and headcount tiers are preserved without exposing exact figures.
The file you're asked to share is usually the one with the most sensitive content. These are the situations where anonymization is the only safe path.
The agency needs the data structure to build their deliverable, not the real contact details. Replace names and emails with fake ones and send them a file that looks real enough to work with.
Your dev or QA team needs realistic data to test an import flow. Export production records, anonymize every PII column, and hand over a file that behaves like real data without exposing any of it.
A regulation or internal policy requires that data stored in a non-primary system be anonymized. Run the export through the anonymizer before loading it into your data warehouse, BI tool, or analytics platform.
You're working with a contractor who needs a dataset to do their job but doesn't have a business need to see real customer data. Hash IDs, fake out contact fields, generalize deal values — send them a file they can work with, not a liability.
The Data Anonymizer runs entirely in your browser. No server receives your file, no third-party API processes your rows, and nothing is stored anywhere outside your own machine.
Every step — parsing, applying strategies, generating fake values, building the output file — happens locally using browser APIs. The download is created as a Blob in memory and handed directly to you without touching a network connection.
Anonymization configs you choose to save are stored in your Asphorem account so you can reuse them. The file itself is never saved.
No. The Data Anonymizer is entirely client-side. Your CSV is parsed and processed in your browser. No rows, values, or file metadata are transmitted to Asphorem or any third party at any point in the process.
Redact replaces every value in the column with [REDACTED]. The original value is completely gone. Use this when you want to suppress a field entirely — internal notes, free-text comments, or any column the recipient doesn't need.
Hash computes a SHA-256 fingerprint of each value. The result looks like gibberish but it's deterministic: the same input always produces the same hash. This means two anonymized files can still be joined on a hashed ID column without ever exposing the real IDs.
Yes. After processing a file you can save the per-column strategy configuration as a named config. The next time you upload a file with the same column structure, Asphorem detects the match and offers to apply your saved config automatically.
Free accounts can save up to 5 configs. Pro accounts get 50.
Files up to 50 MB are supported. The free plan processes up to 5,000 rows per file. Pro extends this to 150,000 rows. Because processing happens in your browser, very large files may take a moment — a progress bar shows how many rows have been completed.
Fake values are generated by a faker library running locally in your browser. The output looks plausible — realistic names, properly formatted email addresses, phone numbers that follow the right pattern — without corresponding to any real person. It's realistic enough for testing and sharing without being traceable.
That depends on the strategies you apply and how a regulator would assess re-identification risk for your specific dataset. The Data Anonymizer gives you the tools to remove or obscure personal data, but it does not make legal determinations. If compliance is a formal requirement, review the output with your legal or DPO team.
Free plan included. No credit card required.
Start for free →