AI transparency
How your AI employee works
Transparency about automation helps you and your customers trust every conversation.
What the AI knows
Your AI employee uses information you provide: website content, services, hours, FAQs, tone, and custom instructions. It does not invent prices or policies you have not configured.
What the AI does
It answers customer questions, captures booking intent, and notifies you when a human should step in. On WhatsApp and your hosted assistant link, replies are automated unless you take over.
Guardrails
When confidence is low, the AI asks clarifying questions or escalates to you. You can update answers in plain language at any time from your dashboard.
Human oversight
You remain the business owner. Review conversations, adjust knowledge, and connect optional channels (Google, WhatsApp) on your schedule.
Limitations
The AI is not a lawyer, doctor, or emergency service. It should not provide regulated professional advice. Configure handoff paths for sensitive topics.
How AI answers
The model generates replies from your configured knowledge base. It does not browse the web or invent prices, policies, or availability you have not provided.
Human override
You can disable AI on any conversation and reply as yourself. Owner notifications alert you when booking intent or escalation triggers fire.
Data usage
Conversation content is used to serve your account — not sold for advertising. See Privacy Policy and Data Processing for retention and rights.
Hallucination policy
When the AI lacks information, it should ask clarifying questions or offer to connect the customer with you — not guess. Report recurring inaccuracies so we can improve knowledge configuration.
Confidence & escalation
Low-confidence topics should escalate to human review. Configure handoff for regulated advice (legal, medical, emergency). The AI is not a substitute for licensed professionals.
Responsible AI principles
Transparency, owner control, proportionate automation, and honest limitations. We do not claim the AI is human.
Also see: AI overview, Acceptable use, Privacy.