VeraZero Anonymization Demo
Ready
Private AI walkthrough

Private by design. Visible in three steps.

Send one message and this page reveals exactly what happens before inference, during placeholder-based processing, and after the final answer is restored for display.

How it works

Sensitive details stay out of the non-private model.

  1. 01

    The user submits raw content, including sensitive data.

  2. 02

    A privacy-protected gateway obfuscates the sensitive values before continuing the workflow.

  3. 03

    The non-privacy-protected model sees placeholders only, effectively anonymizing the request.

  4. 04

    The response is rehydrated for the originating user, so the sensitive values are restored only in the final step.

Explore the full VeraZero product
Single message demo

One prompt. Full privacy trace.

Paste one message to reveal how sensitive details are protected, how placeholders travel through the model, and how the final answer is restored for display.

  1. 01
    Protect

    Original details are hidden before processing.

  2. 02
    Process

    The model works only with placeholders.

  3. 03
    Restore

    The answer is reassembled into readable text.

Press Enter to run the walkthrough. Shift+Enter adds a new line.