Protect sensitive data before AI touches it.
DataShield is the offline gatekeeper: classify columns, mask PII, and create repeatable “trust receipts” so you can safely run AI workflows without shipping raw client data.
What it does
Classify Identifies PII and sensitive fields (emails, phone, SSN-like patterns, identifiers, measures, dates, codes).
Mask Applies offline scrambling so the AI never sees raw values.
Why it’s hard
Teams usually “hope it’s fine” or build one-off masking scripts per dataset. That breaks joinability, loses formats, and leaves audit gaps.
How we make it easy
One consistent workflow: upload → auto-classify → choose strategies → export shielded files + an attestation you can show to security/compliance.
Setup (Desktop)
Install and launch locally. Nothing in this step requires cloud credentials.
How to use (5 minutes)
1) Open a CSV or Excel workbook.
2) Review the auto-classifications (PII / Identifier / Measure / Date / Safe).
3) Enter a passphrase (keeps masking consistent across files, so joins still work).
4) Export the shielded file (safe for AI workflows).
5) Run Workbook Analysis or your orchestrator workflow on the shielded file.