- OverviewUse the assistant
Safety & limits
Built to refuse — ambiguous identity, low confidence or a policy breach gets a clear no, not a guess.
What is it?
The AI Engine is built to refuse. Ambiguous identity, low confidence or a policy-breaching prompt each return a clear, structured refusal with a reason you can act on, rather than a best-guess answer. It never sends comms without approval, never merges guest records on its own, and never crosses between businesses.
Every prompt, draft, approval and refusal lands in an append-only audit log with the time, the person and the model version — entries cannot be edited or deleted, so your compliance team can review a full week of activity without anyone’s help.
The AI will — the AI won’t
- ✓Read your live profiles on every prompt.
- ✓Back every claim with a real detail on the profile.
- ✓Log every prompt, draft, approval and refusal.
- ✓Refuse clearly, with a reason you can act on.
- ✓Record the model version behind any value.
- ✕Send comms without your approval.
- ✕Merge guest records off its own guess.
- ✕Cross between businesses with prompts or data.
- ✕Make up a value to fill a gap — it refuses instead.
- ✕Be overridden by an “advanced mode” toggle.
FAQs
Connected pages
See Safety & limits running on your venues
Thirty minutes is enough. Book a demo or send a question — both routes land on the same team.

