Oracle Gives OPERA Cloud AI Away Free, and the Hotel AI ROI Math Shifts

Oracle made its OPERA Cloud Assistant — AI-assisted room assignments, rate descriptions, multilingual translation and plain-language operational answers — available worldwide to OPERA Cloud customers at no additional cost, resetting the return-on-investment calculation for thousands of properties. Wyndham, which runs more than 2,100 properties on OPERA Cloud, is already putting the assistant to work.

The rollout anchored a show where AI dominated every surface of The Hospitality Show in Miami Beach, November 2–4, from the Visionary Stage sessions to expo booths and the Miami Nights reception. The loudest signal from operators was not enthusiasm but a demand for proof: what is the return on all this investment?

What did the panelists say?

Tuesday morning's session, "AI as Change Agent: What's Next for Travel and Hospitality," was moderated by PwC's Jeanelle Johnson, Managing Partner and Travel Lead Client Partner. Panelists were Benu Aggarwal, Founder and President of Milestone; Peter Chambers, EVP and Chief Information Officer of Montage International; and Priya Chandnani, EVP of Commercial Strategy at Sage Hospitality Group.

Chambers said Montage wants AI for speed and scale, not as a replacement for people. AI optimization, in his view, is essential for getting a hotel discovered, and it can help a hotel understand how a guest's stay went. He saw both sides of the distribution question: AI could be the industry's greatest opportunity, but it could also open the door to entirely new businesses that step in between hotels and their guests.

Aggarwal expects more bookings to move through OTAs, but said hotels that build a guest relationship can still own the direct channel. Her first fix is data consistency, so large language models can find hotels at all. She described one project in which Milestone took every piece of data a brand had collected over years and built a new layer to understand and optimize it.

Chandnani separated the transaction from the stay: whoever handles the booking gets the data, but the hotel owns the stay and the communication around it. Her merchandising pitch was concrete — let guests book experiences, not just rooms, and curate those offers, a spa getaway for example.

What shipped on the expo floor?

The vendor picture showed AI moving from add-on to embedded infrastructure:

• Agilysys has shipped more than 30 AI-powered features and modules under its guestsense.ai strategy, organized around four pillars: multi-modal user experiences, hyper-personalization, agentic process automation and revenue intelligence. Features run inside Agilysys PMS, Book, InfoGenesis POS and Eatec. Two more modules are in development: Revenue Intelligence and a next-generation central reservations system.
• Oracle Hospitality, beyond the free OPERA Cloud Assistant, offers AI Agent Studio for building agents around industry workflows. It has built about 50 agents but acknowledged many hotels still lack a clear idea of which agents they want.
• Canary Technologies announced a global rollout of its AI Voice with Wyndham in December 2025, after a pilot at more than 700 hotels. The tool answers calls, handles booking support and responds to guest questions. The ROI case is blunt: industry data shows hotels miss up to 30% of inbound calls, and about a third of those calls are booking attempts. Canary also launched a Hospitality AI Agent Studio and an Agentic Sales Coordinator that carries group and event leads from first inquiry to confirmed booking.
• Cloudbeds runs on Signals, which it calls hospitality's first foundation AI model, using causal AI — modeling cause and effect rather than correlation — for forecasting, pricing and guest marketing. Ask Signals, a conversational interface for front desk teams, revenue managers and GMs, is in early pilot.
• Hapi works at the data layer, moving data between hotel and provider systems in real time into a single hospitality data model, and has hired a chief product officer with an AI-platform background.

What should hoteliers ask before signing?

Five due-diligence questions emerged from the expo floor:

1. What business problem does this solve, and how will we measure it? Set a baseline and agree on metrics — labor hours saved, direct bookings won, upsell revenue, response times.
2. Is our data ready? Identify which systems the tool draws from, whether data is consistent across properties, and what cleanup or integration work precedes results.
3. Who owns the data and the AI's output? Ask whether guest data will train vendor models, who owns generated content, and what happens to data at contract end.
4. What new security and privacy risks apply? Establish how data is protected and stored, how access is controlled, and how the vendor handles a breach or an AI error that reaches a guest.
5. Who is accountable when the AI gets it wrong? Governance means knowing where people stay in the loop, who reviews the AI, and ensuring the team understands AI well enough to oversee it.

The pressure to adopt is real, but the metrics for measuring AI payback are still being built. HFTP's Hospitality AI Exchange, a new two-day event from HFTP's AI Collective covering operations, finance, governance, IP and security risk, runs February 2–3, 2027, in the greater Dallas area — the next checkpoint for an industry still pricing its AI bets.