Webfold
Menu

Use case

Redact personal data or secrets

Find personal data or credentials. Remove them before you share, index or train.

The problem.

Text leaves your hands for a vector store or a model prompt. It can carry names. It can carry card numbers. It can carry API keys. You want a redacted copy to pass on.

How Webfold helps.

Step 1

Scan as a CI gate

Exit non-zero when anything sensitive is found.

corpusforge scan ./data

Step 2

Redact

Replace every finding with its entity type.

corpusforge redact ./data/notes.txt -o notes.redacted.txt

Step 3

Or call the service

Post text and get redacted text and the findings back.

POST /redact {"text": "..."}

Products used.

PII redaction →

Find personal data or secrets before they spread.

Corpusforge →

From raw sources to a governed training corpus.

OCR →

Read text from scans or image-only PDFs.

Worth knowing.

Note

  • Treat findings as a first pass. Health data needs a person in the loop. So does justice data. So does heritage data.

Try it on your own data.

Run the command-line tool today. Hosted API access is coming.

Join the waitlistRead the quickstart