Back
EP 005
28 Min
Vio.com answers urgent travel cases in as little as one to two minutes. But Maisa Breve, who leads customer experience there, says that when a booking falls apart at midnight, what saves the customer is human judgment, not speed.
Travel is one of the most unforgiving support environments there is. When someone's flight lands in three hours and their hotel reservation is missing, there is no "we'll get back to you tomorrow." Maisa Breve leads customer experience and quality at Vio.com, a travel booking platform serving over 100 million visitors a year, after nearly five years at Royal Caribbean. On this episode of the Fini Podcast, she explained how to deliver empathy through a chat window, where AI earns its place, and why the hardest cases still need a person.
Meet Maisa Breve
Maisa spent nearly five years in guest services and operations at Royal Caribbean International before moving into digital customer experience. At Vio.com she has risen from front office manager to leading customer experience and quality. Her foundation is hospitality, including a background in culinary arts, which shapes a service-first instinct she now has to translate into a chat window and a phone line.
Empathy without body language
Maisa's central challenge is delivering white-glove service without the tools hospitality relies on. In person, you can show a customer you care; online, you have to rebuild trust without body language, tone of voice, or physical presence. That takes greater clarity and what she calls intentional empathy, carried through written words or a phone call. Her team invests heavily in training so agents can convey warmth and credibility in text, and she is honest that some of it has to be innate: if the instinct for people is not already there, training only goes so far. To hit a 10-minute response time (and a one-to-two-minute SLA on critical "I'm at the hotel and there's no room" cases) without losing that empathy, the team leans on staffing, training, and customizable templates that free agents to personalize rather than start from scratch.
Where AI helps, and where she holds it back
AI already adds real value at Vio.com on the first line: a chatbot collects the booking number and the issue, filters urgency, and answers common questions, while suggesting empathetic, personalized phrasing to agents on the harder cases. Where Maisa holds back is anything that needs judgment. AI performs well inside defined rules and standard scenarios, but travel constantly throws cases the team has never seen, which demand flexibility, exceptions, and creativity. She estimates AI could take around 80% of volume, but a meaningful slice will still need a human. Importantly, customers can always choose a person, and the bot hands off the moment someone keeps asking for an agent, so a human delivers the first real response on urgent cases. The thing she most wishes AI could handle but cannot yet: out-of-policy refunds.
Travel is unforgiving: managing the team
The hardest part of building a travel support team, Maisa says, is that mistakes are immediately visible and emotionally charged. A lost booking can wreck plans, money, and a trip someone saved years for, so the job is maintaining consistent quality under pressure, making fast and accurate decisions 24/7, and supporting agents emotionally through it. Her team is fully international, and her advice for that is to stay open-minded, because each culture communicates differently and none of it is personal. She tells the story of a guest from Jordan on a cruise who was upset for days despite her best efforts, until persistent, patient communication broke through and they became friends. To keep agents resourceful, she gives them room to make (minimized) mistakes, backs them with team leads in every time zone, and turns mistakes into coaching for the whole team rather than blame. A clean escalation path and a culture of ownership hold it together.
The metrics that matter
Maisa tracks CSAT first, because it tells her how customers feel about the service and surfaces feedback she can act on. She pairs it with average handle time, distinguishing the time to address the immediate need from the time to fully close the case, since leaving customers hanging drags CSAT down. She also watches public reviews closely, holding a steady 4.3 on Trustpilot, which reflects the whole company, not just support. To keep agents accurate as policies and hotel rules change, a dedicated person updates the knowledge base and the team gets dedicated "catch up" time before each shift to read updates rather than going straight online.
What support leaders should take from this
Engineer empathy into text. Without body language, clarity and intentional warmth are skills you train for, and partly hire for.
Protect speed on the cases that hurt most. Use tighter SLAs for critical situations and softer ones elsewhere, so urgent customers never wait.
Let AI triage, keep humans on judgment. A bot collecting details and filtering urgency is great; out-of-policy and never-seen cases need a person.
Treat international as a skill, not a hurdle. Assume cultural differences, not personal attacks, and let agents learn the customer's style.
Make mistakes coachable. Give agents room to be resourceful with team-lead support, then turn errors into lessons for everyone.
Give agents time to absorb changes. A daily catch-up window keeps a fast-changing knowledge base from breaking quality.
Listen to the full episode
Maisa shares more on building trust online, managing a 24/7 international team, and where she thinks travel support is heading, in the full episode of the Fini Podcast. You can follow her work on LinkedIn.
AI that triages fast and hands off cleanly to a human when judgment is required is what Fini is built for. Book a demo to see it in action.
Leo: Welcome back to the Fini Podcast. I'm Leo. My guest today has spent nearly a decade in high-pressure customer service. She worked at Royal Caribbean International for nearly five years in guest services and operations, and now she leads customer experience and quality at Vio.com, a travel booking platform serving over 100 million visitors a year. Her name is Maisa Breve. What's interesting about her career is the path she took, from hospitality and travel up through operations to leading customer experience at a global digital platform. Travel is one of the most unforgiving support environments: when someone's flight lands in three hours and their hotel reservation is missing, it's fix it now or they're sleeping in the airport. Maisa, welcome to the show.
Maisa Breve: Hi, thank you so much for having me. I'm super honored to be here.
Leo: You spent nearly five years at Royal Caribbean in guest services. What's the biggest difference between managing customer experience in hospitality versus a digital platform like Vio.com?
Maisa Breve: The biggest difference is that face to face, it's easier to communicate with your customer and show them how much you care. Online, building that relationship, interaction, and trust is hard, especially for us as a startup still building everything and needing customers to come back and trust us. So the biggest difference is being able to build the relationship and the trust while providing the same service as if you were face to face.
Leo: You have a hospitality background and studied culinary arts, a service-oriented foundation. Does that mindset translate to digital CX, or is white-glove service harder in a chat window?
Maisa Breve: The hospitality mindset definitely helped me move to digital support, but digital is even more challenging than in person. Real-life service builds strong instincts around empathy, problem solving, and handling emotional situations, which are valuable online too, but harder. In digital you must relearn how to build trust and connection without body language, tone of voice, or physical presence. Creating reassurance, warmth, and credibility through written communication or a phone call requires a different skill set and greater clarity. Your communication needs to be excellent and your empathy intentional, and you have to make white-glove service possible, which is harder to deliver in a chat.
Leo: How do you replicate that sense of care digitally?
Maisa Breve: We heavily invest in training, on the tone you use both in writing and on calls, showing through your voice that you care, that you're human, and that you understand their situation, like being at a hotel at midnight with no room because of an overbooking or a missing booking. We invest in the quality of the staff we hire and in training, and I believe you also need to be born with some of that skill, because if it's not within you, I'm not sure it can be fully learned.
Leo: Vio.com claims a 10-minute response time for check-in-related issues, while many companies are happy with 24 hours. How do you deliver that without sacrificing quality and empathy?
Maisa Breve: We heavily invest in staffing, training, and well-designed templates that agents customize per situation, so they focus on quality and personalization rather than starting from scratch. We also prioritize high-priority cases, like "I'm at the hotel and there's no room," with a one-to-two-minute SLA, while other cases have softer SLAs. We want those critical cases attended to immediately.
Leo: Is AI playing a part in your support today, helping you hit that 10-minute response time?
Maisa Breve: Yes. AI adds strong value in areas like chatbots for the first line of support, getting the booking number and the issue so we can filter urgency, and providing quick, accurate answers to common questions. It also assists our agents by suggesting personalization, even on critical situations, so as you type it suggests how to make a response better and more empathetic. It helps us scale quality and speed without losing the tone we want with clients.
Leo: Where does AI help versus where are you still hesitant to let it run on its own?
Maisa Breve: AI performs well within defined rules, policies, and standard scenarios. But our business often requires going beyond predefined frameworks, we get a lot of cases we've never seen, with no standard. Many situations demand judgment, flexibility, and the ability to make exceptions or be creative. So on critical cases we only use AI to help with tone and personalization, and I think that will remain for a while. For the other scenarios we already use AI for the first line of response and the bots.
Leo: What role do you see AI playing in the next three years? By 2029, could it step in when a customer misses a flight or loses money on a non-refundable booking?
Maisa Breve: Honestly, no. As amazing as AI is, I think it's always going to be a bot with limitations. The out-of-the-box thinking and exceptions are something I believe AI won't handle. I do believe maybe 80% will be taken by AI, but a small percentage will still need human interaction.
Leo: If AI takes 80% of the easy tier-one tickets, what does the future of entry-level support roles look like?
Maisa Breve: For entry-level agents, that's the time to skill up and develop toward a more senior role. I don't believe senior agent roles, with real human empathy, will be replaced. But agents who can only do the very basic are at risk. So now is the time to deep dive, learn from your peers, superiors, and managers, and secure your place, because AI is coming hard. There will always be a place for the good ones.
Leo: What's been the hardest part of building and managing a support team in travel?
Maisa Breve: Mistakes are immediately visible and often emotionally charged. A missed flight, a lost booking, or an overbooking impacts someone's plans, money, trust, and sometimes a dream they saved years for. So the hardest part is maintaining consistent quality under pressure, ensuring fast and accurate decision-making 24/7, and supporting our agents emotionally, because it's not an easy job when they handle stressed or upset customers at high volume. It requires continuous training, strong escalation paths, and a culture of ownership, because even small errors can have outsized financial and reputational consequences.
Leo: How do you determine what gets escalated to agents, especially as you integrate AI?
Maisa Breve: The bot touches everything first, and the customer can still choose to speak with someone. Our agents then make the final decision on whether it's an escalation that needs a higher-level decision. For now the customer always has the option to reach us, and our agents are trained to read between the lines, understand the problem, and decide whether it's urgent and what can be done.
Leo: How do you ensure empathy in that first chatbot response, and do most people in urgent situations just choose a human now that they have the option?
Maisa Breve: The bot asks basic questions, so customers can tell it's a bot, and under stress they usually just keep repeating "agent, agent" or "speak to someone." After a few times the chat is redirected, so the first real response comes from us. Agents can then use AI to build a strong, empathetic response quickly, or write their own. We take advantage of it because it's quicker and the AI-built answers are excellent, which reassures the customer.
Leo: It's interesting you still give the option to escalate to a human. Where do you think the industry is going, will the first touchpoint always be AI?
Maisa Breve: I do think that in a few years AI will be the first response and probably the first agent, even at our company. It's hard to say exactly. We aim to build the best customer support in travel, so we might keep the option of human agents available, even if more limited. Most companies will use AI as the first agent and escalate only serious situations, but for Vio.com I can't promise, because we really aim to beat our competition with the best team.
Leo: What's one thing support leaders need to know about managing fully international, 24/7 operations?
Maisa Breve: I love it. Royal Caribbean was already global, and Vio is the same but online. My tip is to be open-minded: each culture and country is different, some more demanding, some more flexible, some straightforward, some sweeter. It's basic in customer support that it's not personal, and even more so online, where they really don't know you. Be open and learn about the culture so you understand it's just the way they speak. Our team is completely international and I learn every day, which helps me speak better with clients depending on where they're from.
Leo: Have you had a story where cultural differences changed what the customer expected?
Maisa Breve: Back on the cruise ships, I supported a lady from Jordan whose culture was very different from mine, including a much greater emphasis on respect. At first it was so hard to communicate, and she and her husband were very upset even though I was gentle and empathetic. It took about five days for me to understand how to approach them, and by the end we became friends, even had lunch together, and she gave me a pen with her country's flag that I still keep in my office. They didn't want anyone else to handle their case, it had to be me. It was really hard, but it's amazing what you can get through with communication.
Leo: How do you empower your team to be resourceful and creative when there's no defined step-by-step process?
Maisa Breve: We encourage them to think out of the box and give them the liberty to make mistakes, even understanding it might impact the customer. They're supported by team leads per time zone, so they're never alone and can consult before acting. We give budgets and policies so they can make their own decisions on complex cases. Mistakes are allowed and minimized, and when they happen we use them to coach the whole team, without offending anyone. Agents know they won't be dismissed for a mistake, and from mistakes they learn even more.
Leo: Have you had a chatbot hallucinate and give a wrong answer?
Maisa Breve: We haven't, partly because for now the bots handle very simple things, like your booking and your need, with an option to talk to our team. But as we automate more flows in the future, it might happen.
Leo: If you could choose only three metrics for your team, which would you pick?
Maisa Breve: Customer satisfaction first, because it tells us how customers feel about the service and gives us feedback to act on. Average handle time, because people want fast resolution, especially in critical scenarios, and we distinguish addressing the immediate need from the efficiency of fully closing the case. And public reviews, our Trustpilot score is a steady 4.3, which we're very proud of, because it reflects the whole company and product, not just our service.
Leo: As policies and hotel rules change, how do you keep your knowledge base and agents up to date?
Maisa Breve: We have someone dedicated to updating those, and we make announcements without overwhelming the team. We keep team leads on top of each team, and because each lead has only a few agents, people feel cared for and stay aligned. Agents also get "catch up" time before their shift to read updates, so they don't go straight online without knowing what changed.
Leo: Rapid fire. True or false, in three years most travel support will be fully automated.
Maisa Breve: False.
Leo: The one type of ticket you wish AI could handle but it can't yet?
Maisa Breve: Out-of-policy refunds.
Leo: Finish the sentence: the company that wins in travel customer experience isn't the one with the fastest AI, it's the one with the best...
Maisa Breve: Human judgment.
Leo: Maisa, this was really great. Where's the best place for people to follow your work?
Maisa Breve: LinkedIn, under my name, Maisa Breve. Thank you so much for having me, this was an amazing chat.
Leo: Thank you so much for joining us, and for everyone listening, if you want more honest conversations about what it takes to run AI in customer experience, make sure you subscribe to the Fini podcast. We'll see you next time.
Maisa Breve: See you.
Can AI replace human judgment in travel customer support?
Maisa Breve does not think so. AI performs well within defined rules and standard scenarios, but travel constantly produces never-seen cases that need flexibility, exceptions, and creativity. She estimates AI could handle around 80% of volume, while a meaningful slice still requires human judgment, especially out-of-policy decisions.
How do you deliver empathy in digital customer support?
Without body language, tone of voice, or physical presence, you rebuild trust through clarity and intentional empathy in writing or on a call. Maisa's team invests heavily in training so agents convey warmth and credibility in text, and uses customizable templates so they can personalize rather than start from scratch.
How does Vio.com hit a 10-minute response time?
Through heavy staffing, training, and well-designed templates, plus tighter SLAs for critical cases. "I'm at the hotel and there's no room" situations get a one-to-two-minute target, while softer SLAs cover less urgent issues, so the most painful cases are answered immediately.
Which support metrics does Maisa Breve prioritize?
CSAT to understand how customers feel, average handle time split between addressing the immediate need and fully closing the case, and public reviews, where Vio.com holds a steady 4.3 on Trustpilot.
