In June 2026, Oracle announced new AI capabilities inside OPERA Cloud, including an assistant that lets hotel staff work the PMS conversationally. For any property running OPERA Cloud, that announcement changed the question. It is no longer whether AI belongs in your operation. It is which parts Oracle now covers for you, and which parts remain yours to build: above all, the conversation with the guest.
This guide maps the two layers. What OPERA Cloud’s own AI does, where the guest-facing gaps sit, and how hotels connect an AI communication layer to the PMS through OHIP, with the numbers luxury properties are already measuring.
What Oracle shipped in June 2026
Oracle’s announcement introduced OPERA Cloud Assistant, AI capabilities embedded in OPERA Cloud and positioned around helping staff run the property: real-time answers to operational questions, AI-assisted room assignment, and AI-generated rate descriptions, built into the front desk, revenue and configuration screens, as trade coverage from Skift described it. Oracle is shipping these capabilities to OPERA Cloud customers at no additional cost, which treats AI as a standard part of the PMS rather than a premium add-on.
Read the announcement itself and the coverage (linked below) for the current feature list, because Oracle is iterating quickly. The pattern, though, is stable and worth understanding: Oracle’s AI lives inside the PMS and serves the people who operate it.
Where OPERA Cloud already used AI
Two areas predate the 2026 announcement:
- Upselling through Nor1. Oracle’s Nor1 products score and present upgrade offers around booking and check-in, using machine learning on reservation data. If your revenue team runs Nor1, you already use PMS-side AI.
- OHIP as the connective layer. The Oracle Hospitality Integration Platform is not AI, but it is what makes any external intelligence possible: it streams reservation events (booking created, arrival, check-in, check-out, cancellation) to connected systems in real time. Whatever AI a hotel adds on top stands on OHIP. If the acronym is new, start with our plain-language guide: what is OHIP?
The guest-facing gap
Here is the boundary that matters. The OPERA Cloud assistant answers to your staff, inside the PMS, in the PMS’s context. Your guest is somewhere else entirely: on WhatsApp at 23:40, writing in Portuguese, asking whether a late check-out is possible on Sunday.
Three jobs sit outside the PMS’s remit, and they are the jobs guests actually feel:
- First response on messaging channels. WhatsApp, webchat, email, Meta Messenger, Booking.com. The question arrives where the guest lives, not where the PMS lives.
- Language coverage. International properties field questions in dozens of languages, at every hour, from guests who expect an answer in their own.
- Journey timing. The pre-arrival upsell window, the check-in day information, the post-stay review request. These are triggered by reservation events, but delivered as conversation.
None of this is a criticism of OPERA Cloud. It is a division of labour: the PMS runs the property, and a communication layer runs the conversation around it.
How a guest-facing AI layer connects through OHIP
The mechanics are short. OPERA Cloud emits an event, OHIP pushes it to the connected platform within moments, and the guest journey reacts:
- A reservation is created in OPERA. The guest receives a WhatsApp confirmation with their name, stay dates and quick replies: upgrades, spa times, arrival details.
- Arrival approaches. The pre-arrival message opens the upsell window while the guest is planning, not while they are queueing at the desk.
- The guest asks a question. The AI answers on WhatsApp and webchat from the property’s own knowledge, in the guest’s language, at any hour. Staff see every conversation in one inbox, next to the OPERA reservation data.
- A request needs a person. The AI hands the conversation to the team, with the context attached. At St. Regis Le Morne, which deliberately runs every reply through human butlers, chatlyn’s routing keeps the average response under 5 minutes around the clock.
- Check-out lands, or a cancellation does. The journey closes with a review request, or stops entirely. Nothing messages a guest whose booking no longer exists.
The property changes nothing about how it uses OPERA. That is the point of doing this through OHIP rather than around it. The full technical picture is on the Oracle OPERA Cloud integration page.
What the AI does, and what it deliberately does not
Precision here separates serious vendors from optimistic ones, so this is chatlyn’s own line, stated plainly.
The AI answers guest questions on WhatsApp and webchat from the documents you give it, in more than 100 languages, with an average response time of around 3 seconds. It runs around the clock, and it hands the conversation to your team the moment a request needs judgement, empathy or action.
It does not create bookings, quote live rates or availability, or write anything back into OPERA. Reservation data flows one way, from the PMS to the conversation, and your team stays in charge of the PMS. If a vendor tells you their guest-facing AI will operate your OPERA environment for you, ask very detailed questions before you believe it.
What luxury properties measure
Numbers from properties running this layer today, each from a published chatlyn case study:
- IMLAUER Group moved AI resolution from 6% to 93% after switching to a hospitality-trained AI on its guest channels.
- Hotel Sacher handles 70% more routine enquiries through AI, so the team’s time goes to the requests that deserve it.
- Spa Resort Geinberg reaches a 95% open rate on WhatsApp journeys and resolves 86% of enquiries without staff involvement, while running upsell campaigns for spa and dining through the same channel.
- St. Regis Le Morne uses no guest-facing AI at all, by choice, and still answers in under 5 minutes on average because routing and context do the organisational work. The layer is valuable even where the brand voice must stay entirely human.
Is your OPERA Cloud hotel ready for AI?
Readiness is less about the technology and more about the property: how clean the reservation data in OPERA is, which channels guests already use, whether the knowledge the AI needs exists in writing, how the team wants handovers to work, and what you will measure. We built a six-domain scorecard for exactly this assessment, with a maturity model and 20 practical use cases, in the free whitepaper: Is your OPERA Cloud hotel ready for AI?
Frequently asked questions
Does OPERA Cloud have AI built in? Yes. Since June 2026, Oracle ships AI capabilities inside OPERA Cloud, including an assistant for staff workflows, alongside the earlier Nor1 upselling products. These serve the people operating the PMS.
Can AI make bookings in OPERA Cloud? Oracle’s assistant helps staff work with rooms and rates inside the PMS. chatlyn’s guest-facing AI answers questions and hands over to staff; it does not create bookings or write to OPERA. Whoever executes the change, it happens in the PMS, by your team.
How does a guest messaging AI connect to OPERA Cloud? Through OHIP, the Oracle Hospitality Integration Platform. OPERA Cloud pushes reservation events to chatlyn in real time, and those events trigger the guest journey. Each property runs its own integration.
We are still on OPERA 5. Does any of this apply? Yes. chatlyn integrates with OPERA on-premise via secured SFTP, and the same guest journeys continue over OHIP when you migrate to OPERA Cloud.
Which languages does the guest-facing AI cover? More than 100, on WhatsApp and webchat, answering from your property’s own knowledge base.
Where should we start? Two places. Assess readiness with the whitepaper scorecard, and look at how the OPERA Cloud integration would run at your property. Or talk to us and we will walk through both with you.
