Automatically remove PII from free text

Microsoft Presidio is a powerful tool to achive this goal, we could add it to the normalization step to analyse all free text inputs and remove potential PII (added by user itself by accident or by a malevolent user)
https://github.com/microsoft/presidio

The issue does not name files, tests, or a specific normalization entry point. Start by locating the normalization step in the TypeScript monorepo and review Microsoft Presidio's capabilities; done would mean free-text inputs are analyzed and potential PII is removed, including accidental or malicious user-provided data.

Written by the indexing model from the issue text.