New: ChatGPT, Meta, Pinterest and TikTok feeds from the same verified data. Free audit →
Security

Security and data handling

We work with product data, not with your customers' data. We never log in to your store and never ask for access to orders or customers.

What data do we handle?

  • Your Merchant Center feed (its URL and, if protected, its credentials, stored encrypted).
  • The text, tables and image URLs of your public product pages.
  • The values we produce with their source quotes, and a log of every decision.
  • Login email addresses and secure password hashes (argon2).

How we protect it

  • HTTPS only; session cookies are HttpOnly and Secure.
  • Two-factor login (TOTP), user roles (viewer, editor, admin, owner) and login rate limits.
  • Supplemental feeds are served at unguessable URLs, behind a password.
  • Stored credentials are encrypted (Fernet), with the key kept apart from the code.
  • Every change is logged: who, when, what.
  • Daily database backups; every feed version can be restored.

AI and your data

Product page text is processed by Anthropic's Claude API. We send it no personal data. By default, Anthropic does not use data sent through its API to train models.

Subprocessors

ProviderPurposeLocation
RailwayHosting and databaseEU, Amsterdam (Netherlands)
AnthropicAI processing (product data)USA

If you cancel

You can download every value we produced as CSV or JSON, with the source quotes. On request, we delete all data belonging to your feed.

Questions?

For security questions or to report a vulnerability, write to us; contact details are in the privacy notice.