How to add guest-facing AI to your Mews PMS (and what Mews can't do natively)

Mews is one of the most advanced cloud-native PMS platforms in hospitality right now. It handles reservations, payments, housekeeping workflows, and revenue management with impressive automation. But when it comes to guest-facing AI, the kind that talks to your guests across WhatsApp, voice, and webchat in their own language, Mews has clear functional limits. This guide breaks down exactly what Mews does natively with AI, where the gaps sit, and how hotels are filling them with a dedicated AI intelligence layer, like Vertize, that connects through the Mews Connector API.

What does Mews already do well with AI?

Mews offers strong native automation for back-of-house operations and revenue optimization. Its AI capabilities are real, but they're designed to support your staff, not to interact directly with your guests at scale.

Mews raised $300 million in January 2026 specifically for "agentic AI" development, and was named AAHOA's official PMS in February 2026, representing around 60% of US hotels. The platform now serves over 15,000 customers across 85 countries with a $2.5 billion valuation. That kind of growth doesn't happen without a strong product foundation.

Here's where Mews delivers natively:

• Smart Tips for front desk staff. This is Mews's most visible AI feature. Smart Tips analyzes reservation notes, guest history, external review data, and past preferences to surface actionable insights the moment a guest checks in. If a returning guest had a noise complaint last visit, your receptionist sees that immediately. It's a genuinely useful tool, but it's assistive. It helps your staff make better decisions. It doesn't replace the need for staff to be present and available.

• Atomize for AI-powered revenue management. Mews acquired Atomize in late 2024, adding machine learning-driven dynamic pricing and demand forecasting up to two years ahead. Atomize continuously adjusts room rates based on market conditions, competitor pricing, and demand patterns. For revenue optimization, this is serious technology.

• ADA digital assistant for customer support. ADA handles internal support queries for Mews customers (the hoteliers themselves, not their guests) and reportedly resolves around 90% of customer queries automatically. It's effective for platform support, but it's not a guest-facing tool.

• Automated operational workflows. Mews automates payment processing, sends confirmation emails, triggers SMS check-in reminders, and manages housekeeping task assignments. On the Professional and Enterprise tiers, these workflows become quite sophisticated with the Advanced Guest Experience module, which includes digital key support and basic personalized communication.

All of this is genuinely valuable. Mews gives hotels a modern, well-architected operational backbone. But there's a pattern here worth noticing. Every one of these features operates behind the scenes. Smart Tips helps your receptionist. Atomize optimizes your pricing. ADA supports your team's use of the platform. None of them sit in front of your guest, having a conversation, solving a problem, or closing a sale.

Where does Mews fall short on guest-facing AI?

Every native AI feature in Mews is designed to work behind the scenes or to assist staff. None of them hold a real-time, multi-turn conversation with a guest across the channels guests actually prefer.

This isn't a criticism of Mews. It's a design choice. Mews is built as an operational platform, a central nervous system for hotel data and workflows. That's what a PMS should be. But guest-facing conversational AI is a fundamentally different problem that requires different technology: natural language understanding, intent recognition, omnichannel message routing, multilingual support, and the ability to take autonomous action based on what a guest says.

Here are the specific gaps:

• No conversational AI on guest-preferred channels. Mews can send automated emails and SMS notifications, but it can't hold a back-and-forth conversation with a guest on WhatsApp, Facebook Messenger, or webchat. Considering that WhatsApp alone drives engagement rates above 80% in hospitality, this is a significant gap for hotels that want to meet guests where they already are.

• No voice AI for phone calls. Hotels still receive a huge volume of phone calls, often repetitive questions about check-in times, parking, restaurant hours, and directions. Mews doesn't offer an AI voice agent that can handle these calls. Every call still requires a staff member to pick up the phone.

• No real-time intent recognition. When a guest sends a message saying "the lamp in my room is broken," Mews can't parse that, understand it as a maintenance request, create a task for the maintenance team, and confirm back to the guest. A staff member has to read it, interpret it, and manually log the task in the system.

• No proactive, personalized upselling through conversation. Mews supports manual upselling at the front desk, and Atomize handles pricing. But neither can send a guest a WhatsApp message the evening before checkout offering a late departure for a specific price based on their booking history and room availability, and then automatically process the order if they accept.

• No multilingual guest interaction at scale. Hotels in tourist destinations serve guests who speak dozens of languages. Mews's native communication tools don't offer real-time translation or multilingual conversational support.

• Limited semantic understanding. The native messaging capabilities in Mews are template-based and trigger-driven. They respond to predefined events (a booking is made, a check-in is completed) rather than understanding the meaning behind what a guest is saying.

These gaps are exactly why providers like Vertize exist. Rather than competing with what Mews does well, a guest-facing AI layer complements the PMS by adding the conversational intelligence, omnichannel reach, and autonomous action capability that Mews was never designed to provide natively.

What's the difference between native PMS automation and agentic AI?

Native PMS automation responds to predefined triggers with predefined actions. Agentic AI understands context, makes decisions, and takes action across systems autonomously. The distinction matters because it determines what kind of guest experience your hotel can actually deliver.

Think of it this way. Mews automation is like a well-programmed assembly line: when event X happens, do action Y. Guest books a room, send confirmation email. Guest checks in, assign housekeeping task. These are valuable workflows, but they can't adapt to unexpected situations or nuanced requests.

Agentic AI operates more like a skilled concierge who has access to all your hotel's systems. This is the approach that solutions like Vertize take. A guest asks on WhatsApp whether they can get a room with a better view and a late checkout. The AI checks availability in the PMS, identifies upgrade options, calculates the price, offers the guest a specific option, and processes the entire transaction if they agree. It does this in the guest's language, on their preferred channel, in seconds.

The technical difference comes down to a few key capabilities. Agentic AI uses semantic understanding (it interprets what a guest means, not just what keywords they used). It maintains conversational context across multiple messages. It can chain together multiple system actions in a single interaction: look up the guest profile, check room availability, add a product order, and update the reservation. And it knows when to escalate to a human, using sentiment detection and complexity thresholds to route conversations to staff when empathy or judgment is needed.

Mews is aware of this gap. Their $300 million raise explicitly targets agentic AI development. But for hotels that need these capabilities today, the open architecture of Mews makes it possible to add them through integration rather than waiting for native development.

How does a guest-facing AI layer connect to Mews technically?

The Mews Connector API enables deep, bidirectional integration between external AI systems and the Mews platform. This means a dedicated AI layer, like Vertize, can both read data from Mews and write actions back into it, keeping your PMS as the single source of truth.

The integration architecture works through several mechanisms:

• Authentication and access control. The Connector API uses a dual-token system. The integration partner uses a Client Token to identify the application, while each hotel property provides an Access Token specific to their account. This ensures the AI layer only accesses data it's authorized to see.

• Webhooks for real-time responsiveness. Instead of constantly polling the API for updates (which would be slow and resource-heavy), Mews sends push notifications to the AI layer whenever a relevant event occurs. A new booking, a check-in, a room status change: these events trigger immediately, allowing the AI to respond proactively rather than reactively.

• Data mapping between systems. Product IDs, service IDs, and room categories need to be mapped between the AI layer and Mews. When a guest accepts an upsell offer through the AI, the system translates that into the correct financial line item in Mews's accounting structure.

The specific API endpoints that power guest-facing AI include retrieving customer profiles and preferences for personalization, adding product orders for upsell processing, creating internal tasks for housekeeping and maintenance requests, and checking service availability for real-time confirmation of upgrades or additional services.

This API-native approach means the AI doesn't work alongside Mews as a separate tool. It functions as an extension of it. Every interaction Vertize's AI has with a guest is reflected in Mews in real time, so your staff always sees the full picture.

Hotels on the Mews Professional or Enterprise tier get the most from this integration, since these plans offer unlimited marketplace integrations and include the Advanced Guest Experience module that provides the richest data for AI personalization.

What can guest-facing AI actually do through your Mews PMS?

When a dedicated AI intelligence layer like Vertize connects to Mews through the Connector API, it unlocks use cases that go well beyond what either system can do alone. Here are the scenarios hotels are implementing right now.

• Fully digital pre-arrival and check-in. Twenty-four hours before arrival, Vertize sends the guest a personalized WhatsApp message with a check-in link generated through Mews. The guest completes their details on their phone. Once Mews confirms the check-in, the AI triggers a digital key through integrations like Salto or Apple Wallet. Hotels using this workflow report dramatic reductions in front desk phone calls and the near-complete elimination of check-in queues.

• Conversational upselling that actually converts. Vertize analyzes the guest's profile in Mews, identifies patterns (this guest frequently checks out late, for example), and sends a targeted offer at the right moment. "Would you like to sleep in tomorrow? A late checkout until 14:00 is available for €25." If the guest accepts, the room status updates in Mews and the charge is added to their folio automatically. Hotels report meaningful incremental revenue per stay through automated upselling on channels like WhatsApp, because the AI is consistent and tireless in a way that busy front desk staff simply cannot be. The key difference from traditional upselling is timing and personalization. A front desk agent might remember to offer a room upgrade during check-in, but they rarely follow up with a breakfast offer on the morning of day two, or a spa treatment suggestion when the weather forecast shows rain. AI does this systematically, for every guest, on every stay.

• Instant housekeeping and maintenance routing. A guest messages "I need extra towels in room 412." Vertize identifies the guest, confirms their room assignment through Mews, creates a housekeeping task in the Mews task module, assigns it to the relevant team, and confirms to the guest that it's on the way. Response time drops from minutes to seconds, and staff can focus on execution rather than coordination.

• 24/7 multilingual guest support. With Vertize, guests can ask questions about hotel facilities, request restaurant recommendations, inquire about checkout procedures, or report issues in over 50 languages, at any hour. The AI handles routine queries autonomously and routes complex or emotionally sensitive conversations to human staff with full context, so the guest never has to repeat themselves.

• AI voice agents for incoming calls. Phone calls to the front desk are answered by a voice AI with natural intonation. Vertize's voice agents handle FAQs, take reservation requests, and provide directions or facility information. Your reception team stays focused on the guests standing in front of them instead of being pulled away by the phone.

How should you get started with AI on Mews?

Start with a single, high-impact use case rather than trying to implement everything at once. Pre-arrival messaging and digital check-in is the most common starting point because it delivers measurable results quickly and has a straightforward integration path.

Before connecting any AI tool, prepare your Mews environment:

• Clean your data first. AI is only as good as the data it works with, and industry research consistently shows that only about one in three hotel operators trust the quality of their current system data. Make sure room descriptions, service listings, and product catalogs in Mews are accurate and structured. Convert free-text fields into structured data wherever possible.

• Audit your current integrations. Map out every system connected to your Mews instance (POS, CRM, channel manager, door locks) to identify where data flows and where silos exist. The more connected your ecosystem, the more powerful your AI layer becomes.

• Start with a focused pilot. Pick one use case, one property if you're a group, and one or two channels. Solutions like Vertize typically begin with pre-arrival WhatsApp messaging as the first integration point, then expand from there. Measure the impact over 30 to 90 days before expanding. Track specific metrics: response time to guest requests, upsell conversion rates, front desk call volume, and guest satisfaction scores.

• Train your team on the handoff. The best AI implementations use a hybrid model where AI handles routine interactions and escalates to humans for complex or emotionally sensitive situations. Your staff needs to understand when and how the AI will route conversations to them, and they need to see the AI as a digital colleague rather than a threat. Vertize's handoff system uses sentiment detection to identify frustrated or upset guests and transfer those conversations to a human with the full chat history attached. This means the guest never has to repeat themselves, and the staff member walks into the conversation fully informed. VIP guests can also be flagged in Mews so the AI automatically routes their requests to a dedicated team member for a high-touch experience.

• Ensure your website is ready for AI discovery. As AI-powered travel planners like ChatGPT and Perplexity increasingly influence where guests book, your hotel's online content needs to be structured for AI readability. Make sure your robots.txt allows GPTBot, PerplexityBot, and ClaudeBot so these systems can surface your property to potential guests.

Frequently asked questions

Does Mews have a built-in AI chatbot for guests?

Mews offers ADA for customer support (helping hoteliers use the platform), but it does not include a native AI chatbot or concierge that interacts directly with hotel guests across channels like WhatsApp, webchat, or voice.

Which Mews plan do I need for AI integration?

Mews Professional and Enterprise plans support unlimited marketplace integrations and include the Advanced Guest Experience module. The Essentials plan is limited to eight integrations, which may restrict your options.

Can guest-facing AI access real-time data from Mews?

Yes. The Mews Connector API supports webhooks that push real-time event notifications (new bookings, check-ins, room status changes) to connected AI systems, enabling immediate and accurate guest responses.

Will AI replace my front desk staff?

No. The most effective implementations use AI to handle repetitive, high-volume interactions (FAQs, check-in links, simple requests) so staff can focus their time and empathy on complex guest needs and face-to-face hospitality.

How long does it take to integrate a guest-facing AI layer with Mews?

With a solution like Vertize, most hotels can launch a pilot use case (such as pre-arrival messaging) within two to four weeks. Full implementation across multiple channels and use cases typically takes 60 to 90 days, depending on the complexity of your tech stack.

Does the AI work in multiple languages?

Dedicated AI intelligence layers like Vertize support over 50 languages natively, which is especially valuable for hotels in international destinations where guests arrive speaking a wide range of languages.