The Rise of Agentic AI in Hospitality with Tim Major (Operto)

SPEAKER_01 0:02

Welcome back to Pill Talk Sessions with your host, Jessica Gilman. Today, Jessica is joined by Tim Major, CEO at Aperto, to explore how Agentic AI is starting to reshape hotel operations and the future of the hospitality workforce. They'll discuss where AI is already delivering value, why structured data is critical for success, and how a hybrid model of humans and AI agents can unlock everything from smarter direct bookings to a truly personalized guest experience.

SPEAKER_00 0:31

Tim, thank you so much for joining me for a Pillow Talk Sessions conversation. I'm really looking forward to talking to you. And I'm really especially looking forward to really talking about agentic AI and how that looks in hotels and hospitality, but also what we're gonna see coming up. Because I know that you recently launched with Aperto One and an agentic workforce for the hotel sector. So let me begin by actually asking what that is. What does that look like?

SPEAKER_02 1:06

Thanks, Jessica. Uh, appreciate you having me. It's uh interesting times to say the least with AI and hospitality. Um, Aperto One at its core is an agentic workforce of teammates that help hotels uh assist them in their roles. Uh, really fundamentally, we looked at 18 months ago, we looked at, you know, how to what are common operating models for hotels? What are their org designs look like? What are the different roles? What are the responsibilities? And then we broke down how can AI actually assist in these roles and make either improved uh where there's a lot of repetitive tasks, or looking at where could actually do the work on the behalf of the hotel? So that's really at the heart of what Operto uh one is focused on.

SPEAKER_00 1:49

So when you did that initial thinking, did you you I'm presuming you looked at all aspects of operations, delivery of customers, experience and service, marketing as well. So how did you kind of go about that? And and where have you sort of see, where are you seeing the most opportunity for lift?

SPEAKER_02 2:11

Sure. Um, you know, it's but it was an interesting journey because at first we started where most companies started. We started with chatbots because that seemed like one of the obvious places. And I think as technologists who are working with hotels, but definitely leaning on the technology side, we thought, you know, it's let's start to do, let's chat with guests. That makes a lot of sense. Um, but we found that that's one of the most dangerous areas to use AI initially. Um, and not to say that it can't do a good job of that eventually, but you're you're putting what is perhaps one of the most important engagements you're having as a hotel working with guests, and you're putting the hands of AI at the beginning meant that some conversations weren't perfectly accurate. Um, and a lot of hotels, they really wanted to shine in that guest experience moment. So that they didn't want to use technology in that point. So we we went down that road. Um, and then we decided, okay, let's let's build this out further because because we're seeing that there's possibilities there, but it's not really where the main pains are or where we think AI can do the best job. So we started to build different technologies looking at the direct booking problem. Now, the direct booking problem is not a new problem. This has been around since you know the late 2000s when the OTAs really started to dominate. And we started to look into that. And what became clear is that, you know, where you can know all of the information, AI can play a really strong role. Or in fact, where computers or AI can actually know more than a human, they can even assist in ways that a human can't even perform. So to give you an example on the direct booking side, is if you think about all the data that's knowable, you know, you know your search engine position, for example. It's a seeable, knowable thing. You can know your rankings, you can get that information from seeing it on the screen or pulling it from APIs. You can see that there's where Expedia is placing, where booking.com is placing, where predatory OTAs are placing. This is all information that is knowable. That's actually a computer-based function, like the algorithms that are running Google, that is computer-based to begin with. So you can see all that information, and then you can you can see the data that flows to your website. You can see the intent, you can see the IP addresses, you can see all this information. So we thought this is a really obvious area that AI can dig into, especially agentic AI, because it's not just requiring behind the scenes API programmatic um knowledge. It's actually able to function like a person and go and make some of these more like a lot of the systems that we built are then built, the UXs have been built for people. So agentic AI is actually the perfect system to go and use systems built for people like it's a person and build that further. So that's why we really started to lean into direct bookings.

SPEAKER_00 4:49

I I want to touch on something before we get into the direct booking side, but it's such an interesting point that you made is that the the the place that seems the obvious place that a lot of hoteliers or operators have kind of plumbed for first. It's also what a lot of suppliers are selling, is that guest communication bit. But actually, that, as you say, is is the least good place to place AI or certainly 100% AI, because that's the relationship with the guest. And that's where it can go wrong, is if you don't have that bit right. It's a really, you know, a good observation to make because it is. It's like everyone's selling AI to improve your guest communication. There's the fact that you can use different languages and you can sort of um speed up 90% of what might be really obvious questions, but you do need like that bit also is the is the important kind of relational layer that you would have.

SPEAKER_02 5:50

And I think it's not that it's not that it won't be valuable. I think it just depends on where you're at as a hotel. If this is day one in AI, it doesn't make sense to start there. But if you've been doing AI now for 18 months and you've built up the knowledge and the context, I mean, AI at its heart is all about structured data. The agents need context. So if you've built up a repository that already has hotel data, guest data, you've trained it, you've like trained it not in the LLM sense, but trained it in just watching it, giving it more information, helping it better getting better at its answers, then I think that front-to-house activity, there could be value. It's just if you haven't done that yet, you don't necessarily want to jump into the deep end um as your first step.

SPEAKER_00 6:35

So it's interesting when you're talking about it, Tim, you're talking about it as if it's a person there. You need to train it, you need to develop it, you need to give it all the tools, you need to make sure it's got everything it needs to be a success in its job.

SPEAKER_02 6:47

Yeah. It's it's it's quite fascinating in that. I mean, I came out of the property management system space and and and then coming into the AI space in the last year has I found the more I treat it and think of it like a person, the better the decisions are around how to build around it. And that's why I think it's a very different way of building software in that you're not building for just tasks, a lot of software. You're building for a job or a task where where somebody is then going in and interacting with it. You're actually thinking backwards from what would be the outcome we'd want to achieve and where can AI safely take this on to get to the outcome. Um, and what I find so fun, it would, it would be like, you know, if somebody, so as as the CEO, if somebody, if an AI agent just emailed me today and said, I've seen five things about your business. I think three of them need improving and I can fix two. That would be spectacular if it could do a good job of that. And that doesn't really exist in our work lives. Like there isn't a computer doing that, reaching out to us. We reach out to computers, we turn them on, we go into Chat GPT, but actually having it do the reverse of being proactive. And so building software with that that outcome in mind is really been the complete paradigm shift in how I think about AI and how we build for AI.

SPEAKER_00 8:04

So do you know when and how that paradigm shift came for you personally?

SPEAKER_02 8:09

Um, you know, personally, it's kind of happened in little tiny increments. So about probably about a year and a half ago, I started building um building with Cursor, which is one of the programming platforms that that is really built fundamentally using AI. And I had some ideas for property management systems at the time where I was thinking about booking engines. I was thinking about how booking engines could adapt to the user, like depending on the user that came, the actual UX could change. And so I started just building these things on the evenings and the weekends, and I realized I could prototype an idea in 60 minutes, 120 minutes. And I'll never forget the moment where I built something and it was working and it was functioning better than a booking engine that you know we had built over a number of years in the property management system world. And I remember calling my wife over and saying, I need you to look at this and see that this is not like anything we've ever seen before. This can happen faster than we've ever imagined. And that that I'll never forget that because it felt like the internet. It felt like the dawn of the internet where you know I date myself, but going back to like a 14.4 baud modem and getting that for Christmas and plugging it in and connecting for the first time and realizing that there is this other world of internet places you can access and discover and explore. And I knew that that just felt like that's never gonna be the same. That's gonna be a different day in business. And and when I had this moment with Cursor, it felt the exact same. And I knew that at that point going forward, I wanted to work with hotels to bring this technology in. Uh, and the re the reason for that is because I, you know, having worked with hotels and realized how hard they work, um, how intensely they work, that hotels are this unique 24-7 beast that is is really complicated. And it's always been a very fragmented environment of data systems not necessarily connecting. And it's always been a very challenging technology environment. So when I had this realization with AI, it I just couldn't not move forward working in AI. I knew that there was such a beautiful opportunity here to work with hotels in partnership to figure out how do we use this technology well and ethically. Um so that's that's kind of what got us started.

SPEAKER_00 10:20

Yeah, so it's interesting that hotels never turn off, do they? Yes, of course, people go to sleep, but they they're always on. Um and people sometimes not you can't compare it to hospitals, but it's that sort of it's still always on, always needs to be functioning at uh 100%. I want to talk, go, go into actually the agentic workforce now, because you're developing this at Aperto, but we're also maybe seeing this also broader as the shift. Like it's you've had that paradigm shift yourself personally, but the industry is having that shift as well, where the and you know, we had a guest a couple of weeks ago on the podcast who is an operator of independent hotels within Europe, and we talked a little bit about his plans for having an agentic workforce alongside his human workforce. So, how do you see that for operators when you know, in time when they're bringing in these different agents for different parts of the business going alongside the their human workforce? Like, how does that what's the layer there? Do you think? Are you seeing them all in departments together or like a layer or what what what are you sort of seeing?

SPEAKER_02 11:36

I mean, there's an it it's really a complete range from hotels that are not sure what to do, not sure where to start, to hotels that are in chains and brands that are jumping in with both feet. Um both feet looks like, like you mentioned, where it's a number of agents operating in different areas of the business, from accounting to front desk to direct booking, to even going so far as to realizing that that that that data set that's allowing the agents to function will also allow robots to eventually function. And robots in maybe a different sense, that it's not necessarily your your your um your Tesla Optimus robot, like a humanoid robot, but robots that are very specific to tasks. So take for example with housekeeping. Um, you know, some brands have talked about the challenges of of heavy lifting in aspects of housekeeping. So there could be robots that are task specific to just do heavy lifting, or just task specific to clean toilets, just bringing in those elements and they're all running on the idea of this agentic OS. I guess, you know, some have called it that, but it's basically it needs information to know what to do and where to go and you know and how to do it correctly and so on. So that's kind of the range. Um, you know, in terms of how I see it, is I don't see this as uh there's a lot of hype right now, like like like all new discontinuous innovations, there's lots of hype around it. But I think there's gonna be a a multi-year arc of slow adoption because I think it needs to be a partnership of the AI companies working with hotels. Um, and that's gonna mean, for example, like for example, you might start using a direct booking agent or a marketing agent or a front of house agent. And then I think hotels might realize that instead of having the staff that we have, we might have one less role, but we might add a role that is an AI uh architect or an AI coordinator. And they're someone who's watching the interactions of the AIs, um, you know, assisting them, taking them online, taking them offline, because you really are treating them like, in a lot of ways, like staff. So I think there's going to be a slow arc to learn how do you structure your organization with these agents involved. Um, I think the mistake will be to think of it like software that you just delegate or turn on and turn off. It's going to be something that evolves. And that also means that you're working with your AI company to say, okay, it's been six months. The agents are working great, but we need different skills or we need different guardrails. Um, especially as like we're seeing now the there's there's the recent models coming out of like Claude, we're starting to get to these more advanced models. And so as those models come out, we often also have to work in partnership with the hotels to go, you know, where are the guardrails? You know, what information do we want to share? Um so I think it's it's gonna work that way. And then over time, you're eventually gonna have basically a dashboard where you can see all of these agents interacting and you choose the level of delegation you want to give them, kind of like you would staff. So, for example, you might have a front desk agent and you might only keep it in co-pilot mode where it's it's being used by your staff to give you accurate information. But you might decide for accounts receivable, for sales agent, for booking agents. They're performing very well. I'm gonna give them a full outcome level of uh and so you're almost think of it like playing with levers or playing with levers on a on a sound control board where it's who do you turn to 10 and who do you turn to two, and then watching the overall outcomes uh for your business, basically.

SPEAKER_00 15:14

So it's like you need HR for agents. I guess to use HR in Canada, human resources. Yeah, we have we have HR. You need that, and company policies. You need company policies.

SPEAKER_02 15:27

AI might need a vacation sometimes, you gotta fool that out with HR, right?

SPEAKER_00 15:34

They won't need is a vacation. Um but I mean it is fascinating, isn't it? And and and the way and and that sort of realizing, because I think a lot of us see, you know, let's just use AI for XYZ, you know, let's go all in, let's go all in and build, you know, build our own agents, etc. Test learn. But actually thinking about it in that really considered way of what guardrails, which elements that they might be safe in. How do we how do we sort of do the HR bit by developing them, by um making sure they stay that they're working well with with the other agents or or the other parts of the business, sunsetting them when it's time to sunset.

SPEAKER_02 16:24

Yep, yep.

SPEAKER_00 16:25

Bringing in those other skills, you know, it's it's actually far, it's actually more complicated, perhaps. You know, we hear a lot about how AI is actually making a lot of us have to work a lot harder and a lot longer because we're all trying to build, you know, new systems and processes. But actually that complexity is it's not suddenly going to be, oh, everything's done by AI. Actually, it needs a lot of oversight and thinking.

SPEAKER_02 16:49

Well, and it needs a lot of it needs structure. So if you think about, like actually, if you think about a different maturity levels of a business, that as a business matures, it develops roles and responsibilities, you have processes in place, you have operating procedures, you have KPIs, leading indicators, lagging indicators, all of that structure will actually be needed to make AI successful because it needs those guardrails and it needs that knowledge. It's not this panacea where you throw it at a random problem and it will suddenly structure your business. If your business is unstructured and kind of disorganized, then AI will be unstructured and disorganized. Um, and that's why I think it's it's dangerous just to tackle AI and throw it into all areas until you have some of that maturity and systems in your business to then inform the agents.

SPEAKER_00 17:38

So we should all ignore the LinkedIn posts that says, you know, your clawed agents will power everything for you. You don't actually have to work, you don't have to hire any staff, you know, all of those things, which may be my LinkedIn is filled with run out.

SPEAKER_02 17:54

Yeah, and I guess my question would be is has that ever happened? Have any of those headlines ever been true? Right. And uh I think AI is the next leap, but it's yeah, it doesn't just take care of all your problems. Hopefully, one day it will. That would be great.

SPEAKER_01 18:06

On the beach, while AI does it, but I think we're a few years away. Just a quick pause in today's conversation. This episode of Pillow Talk Sessions is brought to you by Abode Worldwide, the strategic communications and public relations agency partnering with ambitious hospitality tech brands to build brand authority and create category leaders. From thought leadership to global media relations, Abode helps the companies transforming how we stay, operate, and experience hospitality to tell their story with impact. If you're a tech company shaping the next era of hospitality, learn more at abodeworldwide.com. All right, let's get back to the conversation.

SPEAKER_00 18:40

Bring something that you mentioned, Tim. You to you you just said it quite in passing a few um moments ago, which is ethics. So looking at the ethical use of AI and how that then comes in. And so that will be around data and privacy uh and those things. What what are your views on that? And what is it that we need to be thinking about as an industry around ethics and privacy and um trust and control and and those sorts of things?

SPEAKER_02 19:10

Yeah, it's a it's a complicated question that's still unfolding. But how we think about it at Al Perdo is that if you think about if your end goal as a hotel is to have a great guest experience, there is this temptation to wade into everything you might know about a guest and use it to curate the best guest experience. So that may look like everything they said, everything they've posted on social media, many, many things are knowable now about a person who has put a ton of information out there into the world. It the question becomes is which information should you use to actually make decisions or inform the agent? So how we look at it is that if if an individual has consented to give us information based on, you know, pretty straightforward concepts of privacy policy, um, and that it's been openly disclosed, uh, then we can use that information to help that guest. So for example, if a guest is at a hotel and he's and that that person is telling me things about what they want, what they like, um, they purchase something and that's openly done, we'll use that information. But for example, we won't go to their social media page, scrape all of their social media, and go down the recesses of their Reddit posts and their Twitter posts, and then from that surmise something we think they would like, because we think that's then going too far. And we also think more importantly, it would create really creepy interactions. So if the agent suddenly responds to you in a way saying, you know, hey, I see that you love Star Wars blankets and we've put one in your room, and you knew that you only posted that in 1996 on uh on a Facebook page, that's where it becomes that's not the type of interaction you would ever want to have with a guest. That would break trust. So thinking about those walls, and so those are the types of decisions we make when we're ingesting information, going, should we ingest that? Should we store that? Is that is that ethical to store? And and so we're we're we're kind of erring on the side of of of being cautious um, you know, as we go.

SPEAKER_00 21:11

Yeah, it again, it's it's something that we've we're still coming to figure out, aren't we? All of us, but to have that as a structure for you at Aperto, that these these are the you know, the red lines, and to be thinking that, because maybe hoteliers are not always thinking about that. They're not always thinking um about the the the challenges or the risks, should we say, of using AI and agentic AI. Yes, of course, in a broad stroke way, but not necessarily in the in the more specific use case way.

SPEAKER_02 21:42

I think I think guests are will appreciate hotels that are using data in an ethical way. Um I think, I mean, I we've assumed that is true for the brand promises of of all hotels, um, but but I think guests are going to value it too. Because there could be a road where companies ingest so much information. Information that it becomes some of those lines become blurred.

SPEAKER_00 22:03

Mm-hmm. Can we go back to the the roles within hotels that that might be shifting? So what what kind of human roles and human skills do you think we're going to increasingly need as we build up agentic workforces that go alongside the human workforce?

SPEAKER_02 22:24

I think there's a there's a couple of different roles. Um I think that this idea of how do you structure your hotel into basically a form of data which is like questions and answers, which is what LLMs have come to expect. Where search engine rules were a little bit different. They were, you know, based on keywords and content and sort of that spider of influence that goes out into the internet, where now LLMs are really looking for the structured data. So I think you might need someone in your company to be considering that. Um, you know, how do we write about our hotel? How do we document our hotel? How do we take all the information that's in our PMS and potentially structure it? And while we're looking at that ourselves to do that with technology, I think it's also important that there is a curator of sorts that wears that hat inside the business, knowing that all these agents will be relying on this now and in the future. And eventually it could be that not only are the agents relying on it to work with humans, but agents working with agents could be, you know, working together. So for example, you could have, as individuals, we could eventually have our own travel agents of sorts, which could be a Claude or a Chat GPT, or it could be something that's more bespoke specifically for yourself. And it knows that every year at uh during winter, I take a certain type of vacation with certain types of qualities and it's outlooking. So you need to be able to have curated information to make sure that your agents are available to those agents. So I think that's sort of that content writer or orchestrator or AI curator role could be important. Um I think in terms of looking at roles inside the business, I think wherever there is repetitive tasks that are knowable in terms of process or procedures, could be accounts payable, something in accounting, it could be um certain marketing roles in terms of, especially in terms of digital advertising or direct booking, where any of these areas where things are knowable and steps can be written down in terms of the tasks required to complete an outcome, those tasks can be given to AI. So I mean, I guess that goes to the HR component, but also goes to thinking about which of those roles are in your organization. Like I would, I would think it would it would be like sit down with your sit down with your hotel, put down your org chart, and for each of the roles, write out what tasks are very creative and require almost a um a random thought to connect the dots to to come up with something. So for example, uh something creative might be coming up with a very unique wellness package that is a combination of you know, people in your region, suppliers in your region, or something that you're gonna bring into the hotel, but you also have to think about the experience, you have to think about the emotional side, you have to think about, you know, the environment, you have to think about the architecture. Maybe that's very creative. So that makes sense for, you know, that's where humans can play the strongest role. But if you go to the other end of the spectrum for every role, what what tasks are just redundant or let are not very well informed? So for example, I take, I take like Google rankings or search engine optimization, is that's always been a classic area where stereotypically agencies give you these long reports. You have acronyms, costs per click, and impressions and original users, and there's all these sort of facts that you need to learn as a person where AI can actually just know all of that. There's no need to actually know that there will be certain knowledge that will no longer be required because AI will just go, we know all of how we know how this entire system works, and we will just get you to the outcome. There's no need to really talk about it anymore. What do you want to do? So I would I would take that org chart and look at all those rules and then start to think about each of the agents that are out there, and then let's go, let's try them out. Let's try out an agent in that redundant area in that task. It could be starting with accounts receivable, for example, um, or something along those lines. And and then just get used to the idea. I think AI is one of those things that until you use it, until you use one of these agents and work with it for three to six months, it's really difficult to understand it.

SPEAKER_00 26:27

Uh so that actually brings me to a question about customers or potential customers or just hoteliers in general that you you're speaking with. And I'd love to know the ones that what you what they're saying to you. So I'm sure you're gonna have everything from those that are all in, like you know, I want to do an I want to do it yesterday, to those that will be very hesitant. What what are ho hoteliers and operators kind of feeding back to you? What are you seeing as being the sort of the the optimistics kind of first adopters to the the ones that are very hesitant?

SPEAKER_02 27:06

Yeah, it it's um it's interesting because we're chatting with hotels from 20 rooms to 200 rooms, and there's not necessarily anything consistent even in room size. Um you know, some hotels are they're getting excited about the idea that this technology can tackle some problems that um they didn't think we're able to tackle before. Um and I take the I take the example because a lot of the conversations have been around direct bookings. And and so, you know, we're positing and showing them technology that's going to tackle it differently. And so they get excited about that. I'm finding excitement. I'm finding I'm also finding that the anger still exists and always have for some of the OTA challenges and the commissions that they pay. Um, we've had a lot of hotels share their like their PL statements showing us how much they pay in commission. And they've become complacent about it, almost sad about it. Um, so it's interesting. I've I I actually find these conversations are a bit emotional for the hotels that we talk with, which and it's either it's either complacency and sadness and acceptance of of OTA fees, or kind of a I don't know what to do. Like I don't know how to tackle this anymore. The technology is changing so fast. Um, you know, I don't know where we are on the search results page. Um, I don't know where we are in the LLMs, and I don't know what to do about it. But everyone's talking about AI because I I think Jessica, like you said, everyone's reading the headlines everywhere and they feel this FOMO of being left behind, which I think the truth is they're not left behind at all because this is early days. No, nobody, we have not seen a lot of great examples yet where anyone is running AI exceptionally well. Or they're running something that they think is AI that really isn't AI. They're actually just have a regular piece of software that they use. Where AI, the way we define it, a gentic AI, is it's actually doing the outcome. And that's very different. Um, so I I think that would be the main, the main sentiment. I think 80% of the sentiment is I don't know what to do, I don't know where to start, let's try it. Like we need to try it. And I think there's an optimism that that let's just plug it in and learn together because it's the only way we're gonna get there, right? We have to, we need like hotels need the technologists, but the technologists need the hotels. And so we have to shake hands, sit at the same table, and then work together to get AI working. And I think there's been a real optimism for that. Um, that's actually what's been one of the most exciting things is is how quickly a lot of hoteliers are willing to try um and just be open-minded about it. And so that's kind of been that's that's probably been the majority of the sentiment. Um, the other sentiment that comes up a fair bit is is this thinking that they are already doing well, where the the data that they're seeing from you know whoever they're working with internally or externally is showing them that they're doing well. But then we're finding that the AI agents are looking at the other data that it's getting, and it's actually showing them that they're not doing as well as they think. So I think part of this is trying to find a way to a source of truth. Like what data is actually accurate? What is actually reporting on my direct bookings and my return? Um, that's been sort of some of the main points I've heard.

SPEAKER_00 30:28

Is it that they're not doing as well as they think, or is it that the AI is showing them how they could do better? Do you do you think like I know that's a sort of a nuanced point. It's like they haven't seen opportunity that they could have had.

SPEAKER_02 30:43

It's a bit of yeah, it's a bit of both. It's a bit of both. Where I think it's always been challenging to get marketing data. Marketing's a bit of an art. Um, and so it's hard to at attribute where I'm investing and where I'm getting return. I think AI has created some new opportunities to actually create attribution, is really what it's done, where it can tie more pieces together that were really difficult in the past. Um and I also think that it's able to, because it's basically functioning 24-7 on your behalf, you're right. It can see things and then it can action them quicker. So just like with building software, where now, for example, you know, I don't write an idea down, write a spec, do UX, do programming. It doesn't, it doesn't take six months. All of that can cycle in a week. Where similarly with marketing now, I think AI can see challenges, make changes, A-B test it, find the right conversion, implement it, and it can do it all in 24 hours, where that cycle would have taken three weeks, typically in the past. So I think so suddenly now you also have a faster cycle. So you have a faster feedback loop, which means you're also finding more opportunities faster. Or, or getting to the the cul-de-sac saying that didn't work. We need to try something totally different. Where in the past, that cul-de-sac would have taken six months to get to to go, wait a second, this isn't working. Well, AI is just doing everything much quicker.

SPEAKER_00 32:06

So that's specifically in the marketing agents, the agents that help the direct booking and the marketing. And I guess it's then also looking at front of house potential and back office protect potential as well. But before we sort of talk about that, I want to talk about data and about information. Like how like from your kind of estimate, how prevalent is it that hoteliers actually have the data in the way structured in the way that they need it to be now so that they can be really moving forwards with it with agentic AI for their business? Like how much work is needed still to be done to make sure that that information is there so that the agents can actually do their job?

SPEAKER_02 32:53

I think uh I think a lot of work needs to be done. Um I think where hotels, um, for example, it in the hotels have done a good job in property management systems and in channel managers. They have their rates, they have their photos, they have their descriptions. But in terms of all the information that exists on hotels' websites, what's been fascinating is so so our system will ingest all the information that exists um for a hotel online. Um, we find most of that information is is not accurate and not correct. So, and that's a problem because then the AI agents don't have the right information to actually interact with guests or even help with direct bookings. So I think there's a lot of content work that needs to be done. Um now, there's easier ways to create content than ever before, but there definitely needs to be a focus on it. Um I mean, we found a lot of interactions where where the agents are wrong. It's just because the information the hotel has provided or the website is not correct. Um, and so the the some hotels receive that as, well, AI is failing. Well, actually, AI is not failing. It's just AI is learning and it's just it's not learning correctly. Um so I think there's an interesting big opportunity there to structure all of that data. Um, for example, all the amenities, the surrounding areas of your hotel, all the aspects of your hotel. If, for example, if you could take every single question that a guest has ever asked, have you written the content for the majority of those answers? If you haven't written, found, or you know, put that content somewhere online, then agents will not can't know it. Um, and yeah, I'd say that's a problem. I would say, I mean, if I had to make up a percentage, it's it's maybe 15% of hotels that we've worked with have have that information sorted in advance. Um really, I think the journey of starting with with working with Operto is that so we built this thing called the AI hub that ingests all this information and kind of grades it. Um, it's kind of the first step to to look at your information to see the quality of it. Is this, are we where we need to be? Um and and I think it's it just it's just something that that I I think you you wouldn't know it until you started thinking about AI. It's not obvious until you go down this, go down this road. It wasn't obvious to me that a website needed to have a lot more questions and answers until we started doing LLM reviews. Um so I I don't I wouldn't expect hotels to be very far along yet, but it's something to definitely start thinking about.

SPEAKER_00 35:23

Same with with other elements of the business as well, where agents can come in. Because even if it's in hu even actually if it's in human resources and and in talent and and people development and that kind of thing, you need to have all your data structured in a way that you can you can use an HR agent who can help with development reviews or or help with hiring, you know, those sorts of things. Um there's a lot of work to be done, isn't there?

SPEAKER_02 35:52

There is a lot of work to be done because I mean, for ex yeah, exactly, with writing job descriptions, responsibilities, what would be the preferred outcome of that role if you were to measure every role? Um, yep, there's definitely, definitely a lot of work to be done.

SPEAKER_00 36:06

I feel like I know the answer to this, Tim, but I'm gonna ask it anyway. And so your background is property management software systems. Why why have you come into this new world of agencai AI? What was it that attracted you to the role of CEO of Aperto and to to kind of come into this new development here? Because it is very much at the like we're at the call face, aren't we? We're at the the new frontier here.

SPEAKER_02 36:33

Um I started in property management systems about 15 years ago. And I was at a company called uh RoomKey. And when I came in, I saw that the channel managers weren't always clearly talking to the PMS. There was a delay in data in terms of booking timing from the PMS back to Expedia, and then a hotel would double book. And there was all these sort of little hiccups between all of the providers in the ecosystem. Um, and I thought I was really excited to, and I thought we'll just fix this. Like, you know, we have APIs and asynchronicity and and uh we can, there's an HTNG standard and we can pull all this together, but it was really difficult to do. It's an inherently very complex industry. And when I saw AI start to tackle some of the issues that hotels are having, it it just answered the question that I thought I was gonna solve 15 years ago. That, and I found after 15 years it was just really difficult with pre-existing technologies to do it. So AI just gave me this new invigoration and this new excitement to tackle these problems with new technology. And that and that's really why I made the change. Um, I mean, I'll I'll give you an example where sometimes we'll be, you know, you'll want to interface to a provider or you'll want data to move between the two systems to create a better guest experience, but you can't because either the other partner doesn't want to provide that data to you because it's part of their competitive moat, or they don't have the technical resources. And so maybe now you have to wait until the summer or the fall before you get that data. Well, the difference with AI now is AI can actually act as a human, log into that third-party system as if they're a regular person, go and get the data, record it, um, and then ingest it into our system for use. So suddenly hurdle goes away, and we can actually get to the problem and get to the outcome that I wanted to 15 years ago. Um, and that's what's was so exciting for me to just get to the get to the problem. I guess I'm someone that like when I see a problem, I really, really want to fix that problem. And I don't want to be done until the problem's fixed. And I found that with legacy PMS, we just couldn't fix the problem. We could get most of the way, but we couldn't fix it.

SPEAKER_00 38:50

So Yeah, you're a problem solver. And what better from a problem solver than to be able to to have lots of problems to resolve?

SPEAKER_02 38:58

Exactly, and lots of little buddies now to help me with the problems. Lots of little teammates.

SPEAKER_00 39:03

Perfect, perfect. Tim, to close this out, what my final kind of question to you is future facing a little more. So if we think about the next two, maybe even five years, you can pick the time frame actually. Of what does a hotel with AI agents look like? So what might it look like in workflows and what might it look like from a guest experience perspective, from somebody working with the hotel, you know, in the hotel? What what's what's your sort of vision? And I won't hold you to it because unless we dig this episode out in two or five years. But you know, your best guess at what that might look like.

SPEAKER_02 39:42

Well, I I I I would think of it, I think of it in kind of two areas. One was that area that I just talked about where you had a lot of fragmented pieces. I think we're going to be able to bring a lot of those data sets together to create a much clearer picture of a guest interaction, a much clearer picture of what can we automate with that information. Um, but I think the second half of the thing that gets me excited, which I think is going to be true over the next couple of years, is the concept of hyperpersonalization. Um, and I I relate this to in 1996, I was just starting my first year of university, and the internet was sort of just coming to be. And there was this book written called uh The One-to-One Future. And I remember reading it, and the idea was that technology was going to bring us back to what it was like to live in a small town where if you went to your baker, they knew what you wanted. They knew what you purchased before, and they were gonna greet you. And that was going to be this notion of if you went onto an e-commerce site, that would be the experience. And or if you went to your banker, it would be like, you know, hey Tim, you know, I haven't seen you in two weeks. And it's great to see you. Why don't you sit down and we'll talk about your mortgage? And it was going to be all this personalization, right? Well, I found that hasn't really come true in most of the internet. Um, you have it in pockets, so ads are really good at personalizing, Amazon's pretty good at personalizing, but generally that was never accessible by small to medium-sized businesses, you know, hotels, um, small chains. It was it was just never accessible. So I think AI is going to allow with this structured data this concept of hyper-personalization, where when you come to the website and you come to the booking engine, it will be guided very much to things that are true for you, like authentically true for you as to what you're looking for or what you're hoping to find, as opposed to this generic concept, generic websites combined with, in some cases, generic stays. I think this idea of hyperpersonalization can flow from the digital side all the way through to the experience and stay, and even after you leave. So, for example, after you leave, when do I want to hear from you? What do I want to hear from you? Right? Do I, do I actually want um to a lot of things right now are triggering just straight booking concepts and straight buyback concepts. But I think it can be much more personalized than that to a business traveler, leisure traveler and so on. So I'm excited that there's this opportunity to see hyper-personalization come to life uh in the next three to five years.

SPEAKER_00 42:18

It's providing that brilliant digital guest journey that we are seeing in other areas of our life, like Netflix, even or Amazon, or even on our banking apps and things like that. Um, that great and and building the relationship. It gives that opportunity for a proper relationship with the customer.

SPEAKER_02 42:38

Exactly. And I think I think that's a good thing to think about this is it's not it's not meant to be technology just interfacing with technology. It's really meant to give the hotel all this information about their guests, and technology's helping them with that. So it's not trying to break the personalization at all. It's trying to enhance the personalization. Um and I think that's what's like, for example, imagine having at your fingertips, and it could be wherever you want, it could be on a phone, where when you're interacting with the guests or five minutes before, you know all the things you know to have a very genuine experience through that stay. Um and I think that's just it, I th I'm seeing it already possible now. I think the question becomes is how do we adopt it? The technology is actually becoming the easier part, knock on wood. It's not really easy, but it's becoming the simpler part. It's going to be working with hotels and the adoption of how do we use this well? But I think I think we can get there. I know we can get there.

SPEAKER_00 43:33

And it's actually it is the opportunity to disrupt the travel platforms, the OTAs, in a way that we've never had that opportunity before.

SPEAKER_02 43:43

Exactly. And I think and I think the travel, the the travel platforms have always been the reason customers go there is A, they're aware, but B, they actually provide the experience needed to curate your your trip. Well, if these systems can now help create these hyper-personalized experiences, uh guests will come to expect that from your property and be happy to go there because they might be able to get a better experience than they can get from the OTAs and really flip the equation. So I think that is very possible.

SPEAKER_00 44:12

Absolutely. I know for one, I use OTAs for the user experience. And if that if I'm getting that elsewhere, or that's one of the reasons, if I'm getting that elsewhere, then or a better version of it even. Um that's where as a customer you you go to. Tim, thank you so much for joining me on a Pillow Talk session podcast. I've really enjoyed this conversation, and I think you know, this the A we have to still remember AI agents alongside a normal human workforce, it's actually quite mind-blowing still.

SPEAKER_02 44:46

Yeah, it is.

SPEAKER_00 44:46

So um, it won't be for long, it'll just be completely the norm. Um, but it's still it's pretty mind blowing. Um, but thank you, Tim, so much for for for joining me for this conversation. Thank you.

SPEAKER_01 45:01

Thank you, Jessica. It was a pleasure. That was a lot of fun. Appreciate it. And that wraps it up for today's episode of Pillot Talk Sessions with Jessica Gilliam. Thanks so much for listening. And if you'd like to stay connected and never miss a conversation, head over to Pillotalksessions.com to subscribe. You can also follow us on LinkedIn for more insights and interviews with leaders shaping the future of hotels, short-term rentals, and tech enabled hospitality. Thanks again for tuning in, and we'll see you next time on Pillot Talk Sessions.