What we read and store

Norian connects to your inbox read-only. We never send, delete, or change anything on your behalf. The connection only lets us look.

When a new email comes in from a monitored client, we use AI (OpenAI) to read it and extract a one-line summary of any commitment or request it contains, along with who owes what and roughly when. That summary is what powers your radar and daily digest.

Once the summary is written, we discard the original email. We don't keep a copy of the email body, and we never use your email content to train AI models, ours or anyone else's. What stays in the database is the structured summary, not the source text.

We tell you this at the point you connect an inbox, not just here. Before the connection is made, Norian shows you what it's about to do and links to the Privacy Policy, so you're making an informed decision rather than clicking through a generic consent screen. This is required under the EU AI Act's transparency rules for AI systems that interact with people, and we'd want it that way regardless.

OpenAI processes the email content briefly to produce the summary, under a signed data processing agreement that governs how it can and can't use what it sees. It doesn't retain the email content afterward.

For what happens when Norian isn't sure about something it extracted, and how to correct it when it gets something wrong, see Accuracy and your control.