AI in hospitality — used pragmatically

Artificial intelligence has a hard time in hospitality — rightly so. Too often it is sold as a panacea that in practice creates more problems than it solves. We take a different approach: the AI assistant in Onboard is deliberately small, well defined and always subordinated to the service team.

Concretely that means: AI answers reservation and standard enquiries, drafts response suggestions, spots double bookings and proposes table rotations. The final decision always sits with the team.

Three principles

We follow three simple principles when applying AI to service-adjacent workflows. First: the AI may not send anything that a human hasn't approved. Second: every AI-generated action is traceable in the audit log. Third: guests are transparently informed when responses are AI-assisted.

A good AI is like an experienced maître d' — present, helpful, and never intrusive.

Pilot restaurants report saving between one and three hours of phone and email time per day — without losing the human touch. On the contrary: service teams have more time for the guests in front of them.

In upcoming releases we're extending the assistant with proactive suggestions: no-show forecasts, load-time optimisations and gentle waitlist nursing. Always with the goal of leaving the really important decisions to humans.