Maya / guest22:14
Hi, we’re checking in tonight. I’m still at the airport, but my sister is already there. Can she get in before me?
A late arrival. A conversation that keeps up.
A delayed flight, a changed arrival, a guest with both hands full. See how a voice concierge could turn one interrupted conversation into a coordinated welcome.
Step inside the storyFICTIONAL BUSINESS · ILLUSTRATIVE USE CASE
ONE MOMENT.
A DIFFERENT POSSIBILITY.
Maya is finally leaving the airport. Her sister has arrived ahead of her, the booking is in Maya’s name, and their plans have changed twice. At this fictional boutique hotel, the Prodero agent becomes the first calm moment of the evening.
The guest changes details while the agent is speaking.
Booking information, arrival instructions, and staff tasks live in different systems.
An inviting conversation still needs clear limits around room access and spending.
02 / THE CONNECTION, IN MOTION
Explore a scripted call. Dialogue, timestamps, system results, and hotel actions are illustrative; no call is placed.
A scripted illustration, not a live agent. Select any moment or play the sequence. All people, businesses, messages, documents, and actions in this story are fictional.
Hi, we’re checking in tonight. I’m still at the airport, but my sister is already there. Can she get in before me?
I can help arrange that. First, let’s verify your booking—
Actually—she’s with me now. Sorry! We’ll both arrive together. And could someone help with a cot? I forgot to request one.
Booking verified. Arrival updated to 23:10. One cot is available; staff placement is required. Night entrance instructions retrieved.
I’ve noted your arrival for around 11:10. There’s a cot available, and I can ask the night team to set it up. Would you like that? I can also text you the night entrance directions.
Guest confirmed. Cot request added to the night team queue. Arrival directions queued by SMS. Front desk handoff includes the corrected arrival and outstanding cot setup.
In this imagined workflow, the guest gets a clear plan and the night team gets a useful handoff. An ordinary phone call becomes a connection between the guest’s changing situation and the people who can help.
An imagined outcome to illustrate the workflow—not a customer result or performance guarantee.
We propose GPT-Live 1 for a hospitality conversation where pacing, expression, and interruptions matter. Business reasoning and hotel actions would run through a separately configured backend agent.
OpenAI · GPT-Live 1 model documentation (opens in a new tab)Listening and speaking can overlap, helping the experience accommodate a guest’s mid-sentence correction.
Natural voice delivery is suited to an attentive, welcoming concierge experience.
A separate agent can retrieve booking context and request tools while Live maintains the conversation.
The model is one part.
The business context completes it.
A proposed telephony connection for incoming calls.
Verified booking context and approved arrival updates.
Cot requests and an explicit human handoff.
Guest-approved arrival instructions.
Aster House and this story are fictional. Every conversation and integration action is simulated. Production would require hotel authorization, identity checks, backend tools, and testing; room access and staff completion remain subject to hotel policy.
When plans break, the show goes on.