Skip to main content

Artificial Intelligence, Real Impact


“Nothing will affect the future of humanity more than
digital super-intelligence.” 

– Elon Musk

Machines that can “think” like humans and simulate intelligent’s the stuff of science fiction, but also of your most recent hotel stay or restaurant experience. Chatbots, robotics, voice-enabled devices and other artificial intelligence (AI) -driven technologies are rapidly being leveraged by hotels and restaurants of all sizes to enhance the guest experience. In doing so, they are part of a broader trend. Gartner ( predicts that the adoption of AI and advanced machine learning will be the top technology move in all markets in the foreseeable future.

Hospitality Technology’s own research indicates as much. According the the 2018 Lodging and Restaurant Technology Studies, AI was named a top technology with future-impact for both hotels (70%) and restaurants (38%).

“Even a year ago, most AI activity in the hospitality sector was very much confined to ‘big business’ and the back office,” says Keith Strier, EY global and Americas advisory leader, artificial intelligence, at consulting firm EY ( “However, guest expectations are changing that. Consumers are accustomed to using technology to manage their personal lives and want the same when staying at a hotel or dining out.” 

AI is everywhere and ultimately it is still in rather rudimentary phases. The full power of AI has actually driven even the most science-minded — including Elon Musk and Stephen Hawking — to voice some trepidations. Musk notably was quoted warning of a future world ruled by an “immortal AI dictator.”  

Fears aside, the potential for allowing computers to predict customers’ desires and provide extremely personalized, timely service — is something that the hospitality industry cannot afford to ignore.


AI Everywhere: The Salve for Service

As was noted by Michael I. Jordan, a professor at UC Berkely in an article posted on Medium (, “Most of what is being called AI today is what has been called machine learning (ML) for the past several decades.” Jordan explains ML as “an algorithmic field that blends ideas from statistics, computer science and many other disciplines to design algorithms that process data, make predictions and help make decisions.” 

Newer technologies classified as AI, such as chatbots, have deep ML capabilities. They become smarter with time, assuming more responsibilities, responding to input in individualized fashion, and anticipating guests’ needs. This shortens the wait or eliminates the need for human assistance altogether.

Edwardian Hotels London (, which operates 13 properties in the U.K., harnesses a customer-facing chatbot named Edward that leverages natural language understanding to answer guests’ questions before, during and after their stay. For example: “What time does breakfast start?” and “Is it possible to get extra towels?” Built and trained using a proprietary platform and accessible via guests’ mobile devices, it understands several different languages and also handles myriad requests beyond housekeeping and room service. For instance, it can be used for taxi-booking, and it provides local sightseeing suggestions when queried. It even anticipates returning guests’ needs based on their guest history. Service requests are routed to the proper department, and complaints are escalated to human employees.

To date, Edward has interacted with more than 400,000 Edwardian Hotels London guests, says IT Director Michael Mrini. 

“Giving guests a level of autonomy supported by ‘on-the-ground’ staff, along with a deeply personalized experience,” has led to heightened guest satisfaction and serves as a competitive differentiator for the chain, he notes.

Chatbots are also making their way into restaurants, increasing customer satisfaction by remembering previous orders, making recommendations during the order process, automatically retrieving previous orders, answering questions and more. Like hotel chatbots, restaurant chatbots can be built into operators’ own apps; solutions providers like Concepta ( offer this option. However, some operators are adding a level of flexibility and increasing usage potential by embedding smart messaging into apps and websites with bot-builders and messaging APIs from Oracle (, Gupshup (, and others. 

Domino’s ( last year introduced Dom, a chatbot that takes customers’ orders on Facebook’s messenger app. The chatbot remembers previous orders and tracks current ones. Similarly, Taco Bell’s ( TacoBot provides answers to questions, accepts orders, organizes group orders, offers recommendations and handles payments on the Slack messaging app. TacoBot understands and responds to customer comments made on Slack — sometimes offering “assistance” or silent judgement. One example: customers who mention that they are drunk may receive a cup of water with their order. Some restaurants have gone one step further with chatbots, using them to track loyalty program participants and automatically contact these customers about targeted rewards and


Machine Learning Leading Service

The Sheraton Westport Hotel ( is testing Angie by Angie Hospitality (, a cloud-managed, intelligent digital room assistant with multi-lingual voice and high-resolution touchscreen interfaces. Angie provides guests with access to hotel amenities and services such as in-room dining. Additionally, it enables them to control the room TV, lighting and temperature; listen to music through Bluetooth speakers; and make hands-free telephone calls. Lighting and temperature settings are remembered for future visits.

In addition to improving the overall guest experience, the technology affords the hotel an ancillary benefit: Streamlining guest services and bolstering employee efficiencies frees up staff to focus on more “human” interaction when guests want and/or need it, says Mitch Bolen, general manager, Sheraton Westport Hotel.

Some hotels have begun utilizing facial recognition to identify VIP guests and provide them with special treatment. Meanwhile, on the restaurant side, CaliBurger ( is testing an intelligent self-service kiosk. The system uses AI and NEC ( facial recognition technology to find customers’ previous orders and loyalty/rewards program information, accept orders and process payments.

Oncam ( reveals that hospitality companies can use AI and real-time video analytics to identify and address instances where attention to customers or guests is required. For example, if an AI-based analysis of images captures long lines at a hotel’s front desk, the system can send an alert to a manager along with a video clip, so additional staff can be quickly assigned to the area.

According to Genesys ( and Intelligencia (, operators can leverage AI to analyze guest and customer histories, identifying discrete characteristics in consumer behavior. Armed with this data, they can anticipate and fulfill guests’ needs in advance and avoid alienating guests by offering them irrelevant “perks.”  For instance, if a guest checked out late the last three times he stayed in a hotel and requested a babysitter, AI can anticipate that and automatically offer a late checkout or suggest a nanny before the guest asks. 

A Roadmap to Building an AI Foundation 

Despite its benefits, simply having one or more applications of AI in place does not guarantee success. Sources recommend taking several steps to ensure that implementation does not backfire, alienate guests, spark operating inefficiencies, and place obstacles in the path of maximizing technology capabilities as well as investment. 

Thorough assessment. According to Icebergh (, operators must think about how the AI tool will change the overall guest experience in the short and long term alike and whether the change fits predetermined objectives. Also warranting exploration are whether the tool will place a burden on some guests and/or is too advanced for a particular operator’s customer demographic; whether it will ease common pain points, and if staff will know how to interact with it as well as act on results. If the answer to the last two questions is “no,” a training plan for employees must be devised.

Careful attention to technology selection and integration. Focus on solutions that integrate with other systems such as the PMS, CRM and POS. A lack of integration renders it impossible to offer guests a seamless AI experience and difficult for decision-makers to access the data they need to identify trends and problems as well as formulate new initiatives.

EAST, Miami ( will test Amazon for Business voice-controlled assistants this year. Mihai Bote, CHTP, director of technology, says the number of vendors whose systems are integrated with Alexa played a major role in its selection.

“Other (offerings) may speak more languages, but Alexa has a higher degree of integration — and more skills,” Bote notes.

Proper training of AI solutions is critical. This starts with incorporating as much data as possible, because the more data on which AI solutions are trained, the greater these solutions’ potential to fulfill their promise of a better guest experience. For example, if a restaurant chatbot is not trained to remember customers’ preferences but instead merely accepts orders, it cannot effectively create happier patrons by simplifying order taking. 

“The key here is to think about what you want AI to do, and ensure that its training takes into account whatever data will allow that to happen,”  explains Jim Melvin, CEO of brand and consulting firm Intelligent Transactions (

Whether their interaction with guests is verbal (as with voice-activated assistants) or virtual, proper training of AI solutions also means conditioning them to communicate as humans do, rather than as machines do. 

“Colloquial is best,” asserts Hyoun Park, CEO and principal analyst at technology management consulting firm Amalgam Insights ( “Hotel guests (and restaurant customers) will be very intimidated by an impersonal tone, defeating the purpose of using AI for customer engagement altogether.” 

By the same token, Park says, voice-activated assistants should be able to obey direct and indirect commands, harnessing intuition to do so. For example, if a hotel guest says he is cold, the device should offer to turn up the heat instead of waiting to be told to do so.  

Mrini believes Edward has been so well received by Edwardian Hotels’ guests partly because it has been conditioned not to answer questions or respond to requests with “I don’t know” or “I can’t help you.” Instead, it says, “We’ll get back to you shortly,” then contacts staff to provide assistance. If asked a question to which it does not have the answer, the chatbot sends that inquiry to guest relations and adds the answer to its learning.

Up Hotel Agency ( recommends configuring chatbots to recognize instances where a guest is struggling or becoming frustrated with them. Customers should have an easy way to extricate themselves from automated conversations and connect with a human employee if they wish — for instance, by clicking an icon or using a voice command.

Continuous assessments. Medallia ( advocates evaluating the effectiveness of AI, paying attention to how well the platform is working and whether it is truly enhancing the customer experience, as well as whether the system continues to learn, as
it should. 

This ad will auto-close in 10 seconds