What we see. What we do not.

The app does not create an account. It does not keep the PDF. It does not keep the extracted text. It does not keep the card after you download.

The file is opened in your browser. If you ask a model to write a card, the extracted text — not the file — is sent to a language model on the server and discarded with the request. The API key stays on the server.

The waitlist stores an email address and the time it was added. Nothing else.

A public counter stores how many cards were bound and an estimate of tokens not spent. It does not store filenames, hashes of documents, or any document content.

There is no advertising SDK and no document analytics.