AI Room Service Assistant

The problem

Luxury hospitality is judged on response. A guest wanting extra towels, a dinner reservation or a late checkout expects it handled immediately, and at occupancy the front desk becomes a queue — which is precisely the experience the property is selling against.

How it works

An in-room assistant built on the OpenAI API takes guest requests conversationally, covering dining orders, housekeeping and local recommendations. The integration with Opera PMS is what makes it operational rather than decorative: requests become tasks in the system housekeeping and kitchen already work from, routed to the right team automatically, and the assistant knows the guest's stay context — room, preferences, existing bookings.

A Node.js service sits between the conversational layer and hotel operations, translating between them.

What shaped it

The assistant had to fit staff workflow rather than add a parallel one. A request captured somewhere nobody is looking is worse than a phone call, so everything routes into Opera rather than into a separate queue.

Personalisation also had to stay the right side of comfortable — using stay context to be useful, without making guests feel catalogued.

Outcome

Guest requests were fulfilled faster with less pressure on the front desk, and the experience was personalised to each stay.