This is a public version of the pattern we help organizations build. It is grounded in approved Assisted Intelligence context, it has explicit boundaries around what it may do, it uses deterministic rules when something matters, and it tells you when information is not published.
- It reads an allowlist of our published knowledge. It has no access to our repository, our database, our accounts, or anyone else’s conversation.
- The diagnostic is routed by fixed rules. A model can help you phrase an answer; it cannot choose your pathway.
- Nothing is saved until you fill in a form and press the button. Agreeing in conversation is not consent.
- This conversation is temporary. It is deleted 24 hours after you stop using it, and it is never used to build a profile of you.
Read the privacy policy