How can enterprises become fluent in AI? (Learnings from VOX 2025 conference)

Summary

At VOX 2025, one thing was clear: the best brands aren’t just adopting AI — they’re becoming fluent in it. In this episode of Deep Learning with PolyAI, Michelle Schroeder, our CMO, sits down with Christopher Osborne, our new Chief Product Officer, to unpack what we learned from our annual conference about how the world’s top enterprises are moving from early adoption to true fluency in AI for CX. They reflect on key lessons from VOX 2025, where leaders across hospitality, banking, healthcare, and more shared how they’re using agentic AI to scale exceptional service without losing the factors that make them stand out to their customers. They discuss:

• How Seminole Hard Rock, Boyd Gaming, and Fertitta Entertainment use AI to protect the guest experience, not replace it
• How Metro Bank is turning “a bank in every pocket” into a banker in every pocket
• How DOCS Dermatology, Simplyhealth, and SafeRide Health are expanding access to care through AI-driven patient engagement
• Why the next CX differentiator isn’t adoption — it’s fluency
• How PolyAI helps enterprises balance automation with humanity

Key Takeaways

• The “Agentic Enterprise” is here: Christopher Osborne describes the shift from generative to agentic AI as a decade-long journey — one that mirrors the adoption of cloud computing. The technology now works; the real challenge lies in reshaping people and processes to integrate it into daily workflows.
• AI protects — not replaces — people: Hospitality leaders at VOX, including Seminole Hard Rock and Fertitta Entertainment, emphasized that AI isn’t about cutting staff — it’s about protecting the guest experience and scaling high-quality service across every touchpoint.
• Consistency without sameness: As industries like hospitality and banking adopt AI, differentiation comes from tuning the technology to reflect each brand’s unique voice and personality — ensuring automation feels personal, not generic.
• When AI sets the standard for humans: Nicole Rice of DOCS Dermatology delivered the quote of the event — “We teach our staff to sound as good as our AI.” It captures how far voice AI has come — from a support tool to a benchmark for customer experience excellence.

Transcript

Michelle Schroeder
00:02 – 01:16
Hey, everyone, and welcome to deep learning with PolyAI. So we’re here to help CX leaders get a window into the latest and greatest developments in AI.
So if you’ve been with us for a while, you might remember me. I started this podcast about two years ago with our our CEO and cofounder, Nicola.
But if you’re just joining or if you haven’t seen me in a while, hi. I’m Michelle Schroeder.
I’m Polly CMO. And before we get started, just a reminder to subscribe, give us a like on YouTube, leave us a nice rating on whatever podcast app you’re listening in from.
And now I’m super excited to be welcoming to our podcast today, very special guest. It is our chief product officer, Christopher Osborne.
So Christopher has two decades of experience building and scaling world class product organizations. Most recently, he helped transform DeBell from a 50 person startup into a 3,000,000,000 person plus company.
He’s also led major product organizations at booking. com, served as an advisor to the UK government on digital and data strategy, and now he’s with us helping to build the future of conversational AI.
Welcome to the show, Christopher.

Christopher Osborne
01:16 – 02:35
Well, hello, Michelle, and, thank you for such a kind and generous intro. I think I I think I’m gonna copy paste that, and, and I’m gonna have other people, introduce me the same way because that was really very kind.
But no. Like, so happy to be here.
It’s such an exciting space to work in. I’ve developed a habit, I think, of going to places where they’re at the forefront of what’s going on, but where there’s already or we’re seeing scaled adoption.
And that’s exactly what I saw with PolyAI. Everybody’s talking about agents.
I think if you had a word counter for agentic in most of the, most of the. keynote speeches we’re seeing at big tech conferences now.
It’s off the charts. But I found you guys, and apart from, of course, being attracted to the team and thinking I like these people, The challenge is just so exciting of actually doing it for real, and being in there with big companies and actually having really scale, with the GenTech AI.
So it’s it’s a great place to be, and I’m very happy to be here.

Michelle Schroeder
02:35 – 03:41
I could not agree more. Could not agree more.
And we’re so, so excited and lucky to have you with us and help in helping us scale that. I remember, so I I joined Poly about three years ago, and the big word on the street, your your big kind of word, that got repeated over and over at tech, you know, keynotes then, was generative.
And so, obviously, that’s still very much a part of what we do, but we are just as guilty as every other kind of tech leader in using the word agentic to describe what we do. And so I guess I’ll just kick off, with, you know, what Nikola left us with at our keynote last week at VOX, at our annual event VOX.
He was talking about the agentic enterprise. And just given you’ve got this really kind of perfect dual experience of working directly at an enterprise, booking.
com, like, one of the biggest companies in the world, and, you know, working with us in the forefront of building this for companies like booking. com, I’m just curious, like, what the agentic enterprise means to you.

Christopher Osborne
03:41 – 04:10
By the way, there’s a beautiful edit on YouTube. I think it’s with Sundar at Google IO, where they edit together every time he says the words generative AI, and it goes on for about three minutes.
And I I quite happily shared that with my team before we went on stage with VOX to, to get a bit of competitive spirit, about about the the rise of a Gentec AI. If I think about a Gentec enterprise,.

Michelle Schroeder
04:10 – 04:13
Right.

Christopher Osborne
04:13 – 05:42
I’m quite a simple and straightforward person. And I think the real art of being good at product management is taking something very complicated and being able to make it simple or at least be able to turn it into blocks that are simple to comprehend, and then we can go and build those different blocks and and get to actually a complicated end state.
And I think agentic enterprises are happening. It will evolve over a long period of time.
One of my favorite charts is this, cloud adoption chart, where, if you’re like me, you started tinkering around with AWS 2008, maybe, something something like 2008 was when it first came out. And, of course, you know, if you were a start up person and playing around in your bedroom with tech, you were playing with AWS in 2008.
If we fast forward almost seventeen years and you look at the adoption charts of cloud and AWS, I think we’re over the 50% mark.
Maybe it’s 50 or 60%, and that’s a seventeen year journey.

Michelle Schroeder
05:42 – 05:42
Yeah.

Christopher Osborne
05:42 – 07:37
So, that that shows how long it takes to really gain traction, and adoption and then scale use cases in different enterprises. So.
we’re happening a lot faster, but I also assume we’re on a decade long journey to get there. And the different parts of the journey are really as simple as people, process, and product.
Obviously, I’m in the wheelhouse of an advanced tech company who gets to work with this stuff and is building it and and has a decent idea of what at least what the next two years looks like. I think anybody who’s talking more than two years out is really, really guessing, like, really guessing.
Nobody really knows. And, yeah, it’s as simple as we’re at the stage where we have tech and products which genuinely work.
Right? You can apply them to existing workflows, and you can automate 80 to 90% of it, which for all of us is a huge, huge win in terms of productivity. We’re there with the tech.
In terms of people and process, this is probably the gnarly part right now and and what people what we should pay attention to as somebody who’s producing the technology of how do you get it into the hands of people who are not domain experts and who have not got a PhD, in, you know, in. LLM fine tuning.
And then also how do we actually make it work in a business workflow and process. And, for me, it’s a lot clearer what the product looks like for, you know, for a agentic enterprises.
And we’re now all trying to figure out what it looks like in terms of how do I build this into my workflows, and then how do I have the right people and the right teams, the right processes so we get the most value from this?

Michelle Schroeder
07:37 – 09:24
Absolutely. I I think one of the lucky things about us last week is, you know, obviously, we are the tech company.
You know, you are the the the product leader. You know? And you’re saying the tech is there.
People in process obviously are are still to come, but we had a lot of people on stage with us last week, you know, talking about how they’ve been able to integrate this into their workflows, how they’ve been able to extract a lot more value. And a lot of people shared with us at our annual event last week the things that surprised them, you know, the the the the big successes that sort of came out of adopting this technology and embracing this technology.
And so I just wanna talk a little bit about that. So we had a couple of industry panels.
Our first one being a hospitality panel with Nadir Kim from Seminole Hard Rock, Stephanie Wheeler from Boyd, and Brian Jefferson from Fertitta Entertainment. And these leaders all talked about some of the hurdles that they had to face, you know, adopting this tech and bringing it in and getting people to be excited about AI when AI was being so heavily associated with the notion of replacing people.
And they said, basically, unilaterally, it’s not about replacing people. It’s about protecting the guest experience.
And I really liked something that Nadir said, which is that, you know, his job with this technology is to make the service channel feel like the resort. And then just adding to that, Brian Brian Jefferson said something, which is that, you know, we sell fun.
We’re in the business of selling fun. So how do you do that? You know? How do you think about success in the service channel, success of the technology, you know, really representing the brand and bringing fun and bringing sort of the spirit of the resort? How do you do that from a technology standpoint?

Christopher Osborne
09:24 – 10:43
Yeah. I, on that note, I I worked in hospitality, I guess, or as.
as part of hospitality for six years. And I think it’s a great industry for us to look at and learn from, particularly where we’re working, because for them, the customer experience is what they are selling at the end of the day.
It’s like you are selling an experience in a resort, in a room, in a casino, or in a restaurant. And what I learned from the hospitality business is that they are very passionate about that, and it’s also a key emotional part as well because, if you get that wrong, people will very loudly, talk about this online in the different forums.
And, you know, guess what? It’s just this is your day job is providing this experience, so people really care about it in these industries. So I think it’s.
a very good very good place to start and and to look at. And, yes, we’re fortunate enough to have, some some very big names and some strong customers from from from that group at the event.

Michelle Schroeder
10:43 – 12:14
Yeah. One thing, you know, that was interesting that came up in a couple other sessions is that, you know, AI has given companies like hospitality where they’re competing on experience, they’re competing on how people feel when they’re resorts and I guess the extent of how much.
fun they’re able to bring to people. But just that consistency allows you to scale what’s good at a very, very, like, high degree to to a very high degree.
But what if that consistency also brings sameness to the industry? You know, a lot of folks, especially, you know, you know, not us. But they’re leveraging a lot of the same exact underlying technologies to do what they do.
And so one of the big things that came up was how do you tune this to your brand? How do you make sure that you’re scaling what’s specific and unique to you as a company? And I think what was interesting is that, you know, we had a banking session, like a finserv session, following where, you know, they too, just like hospitality, they’re competing on experience. They’re competing on that personal touch, because that’s that’s really what they have.
You know, financial instruments are what they are. And so we had Metro Bank’s, Lee Prothrow on on our stage with us, and he talked about his vision.
And he said, you know, what we’ve done is put a bank in every customer’s pocket. And what we’ll be able to do soon is put a banker in every customer’s pocket with this technology.
How do you see that kind of proactive relationship, evolving, I guess, across other industries?

Christopher Osborne
12:14 – 13:43
Yeah. I think if I try and generalize it, and by the way, I got excited talking about the hospitality industry and then didn’t quite answer it.
Didn’t quite answer your question, Michelle. So let me let me try and pull that back and then also integrate it, with with what you just asked me.
I think, in general, there’s a couple of points going on here. And in product speak, in CPO or PM world, we’d love to talk about frictionless experiences.
Right? So if your app your you know, let’s say if you are, Metro Bank or or another, you know, relatively modern bank, you’re really and these they actually founded it on the principle of we can give people a much easier and frictionless experience with the mobile app. You know, that was.
the big the big game changer in tech from what? 2010 onwards or something? Something like this. 2000, it was okay.
Well, now we’ve got mobile apps. And that in itself was a revolution in that you could put actually an incredibly powerful computer in somebody’s pocket, and it didn’t feel like a computer.
And, yeah, you know, we put a computer in everybody’s pocket, and everybody forgot it was a computer because it was so easy to use. And then this opened up, you know, a frictionless or let’s say, much less friction like experience than going to a traditional bank.
Let’s just say mobile apps. And then also a new breed of businesses like the Neo banking, fintech experiences where you’ve had, you know, huge players like Metro Bank really come around from people just saying, you know what? We can take a lot of the services you used to have queue up for.
Put that in your pocket. You can do it in a couple of taps.
Already a huge breakthrough. And now, as, was pointed out, we’re opening up a kind of intelligence where you can interact and get much more and get fully professional like customized information, which is highly relevant for you.
And that’s that’s where I think he was going. And the same is for where we’re going with agents.
We are trying to produce, a multimodal experience where whenever you need the support or whenever you need to reach out some contact or just perhaps some answer something, that we’re there. And this opens up and changes the nature of the relationship from fully inbound, reactive, and, frankly, a pain point for a lot of businesses.
Right? Like, if somebody has I’m not a fan of picking up the phone and calling people. I’ve happily got used to texting and using the web for the vast, So I’m not a big fan of picking up the phone.
But if I do pick up the phone, it’s because something is not working or I really need something and, you know, I I’m reaching out. You know, I’m reaching out to that company and saying something is wrong, and I need assistance.
The agents we have now, as of now, really able to work in any channel. We’re working towards things like persisting memory across different channels.
You know, you call, you hit a web form. Perhaps you send an SMS, and it’s not a disjointed experience, but we know the customer who is working with us, the relevant data is in the right place so the agents have the context, and we’re not restarting conversations.
We are continuing an engagement and, you know, hopefully, being part of that relationship with our customer. And, yeah, just like the mobile phone has enabled companies like Metrobank and others to have a much better customer experience, The future and what we can see coming quite rapidly.
And, again, you can have a good idea of what’s coming in the next two years. But unless you are a Steve Jobs, like, nobody’s really a 100% sure how it crystallizes.
But, we have $8,800,000,000, which is probably bigger now, people interfacing with chat GPT. And that’s just one of these services.
So it’s easily 1,000,000,000 people worldwide every week is interfacing multiple times, with a generative AI product to get information, but also now to perform tasks.
But your phone or some other device you have will be able to interact with many different services. Tasks can be performed for you on request, which will become more complicated and and open up quite different way for people to interact, with either services or Indeed brands and and, you know, customer offerings.
And nobody quite knows how it looks like, but the technology we’re building is almost certainly gonna be this part of how a brand can then interface with the consumer world, because undoubtedly, it will be dominated by a few players in the consumer world. Of course, you’ve got the Samsungs, the Apples, and the Googles.
I think OpenAI actually has a very strong chance of being the first new entrant to heavily disrupt the consumer space in terms of hardware and mobile experience. So they’re clearly investing a lot of money there as well as we speak.
Perhaps they make it, perhaps they don’t. The consumer world will be you know, it’ll look fairly similar where these big names own the consumer world, and people have to entrust them to look after their data and to provide them security and privacy and and broker the experience.
But then there will be this agentic interaction where the things that you’re doing on a daily basis or, you know, every hour on your phone. will somehow interface with the different, different companies and different brands that you are that you’re using.
And that’s where the come the customers that we’re working with now, like, you need this kind of interface where people will come to you through this channel. And yeah.

Michelle Schroeder
20:51 – 23:28
Yeah. Enterprises have been adopting two channels basically as long as they’ve been popping up.
I mean, so we see, like, even just from a marketing perspective, you know, as new things arise, whether it’s AR, TikTok, you know, you you figure out how to meet people where they are and where they wanna be. And you talked about earlier sort of the the, you know, the people, the process, the technology.
In that people category is the consumer. And what we’re seeing with that herd of 800,000,000 people interfacing with ChatGPT every single day is this embracing of this technology of conversational AI and, like, the desire to sort of interface with agents.
And so enterprises are in this, like, really, really great position to take advantage of these emerging channels. If you know, we don’t know what it’s gonna be.
Maybe it’s it is like, you know, the assistant, you know, Siri becoming that interface between you and the brand. Who knows? That’s not for us to predict right now.
It’s just for us to be there to design that agentic future and make sure that the enterprise has a place in it. But, you know, just going back to something that you said earlier about, you know, the relevance of the information that the person’s getting, the mobile app, like, the the app in your pocket, the banker in your pocket, back to Metrobank’s, quote.
Really, the theme of all of that, the thing that makes it embraceable technology or an embraceable future is access. And so just, you know, thinking about, patient access and thinking about our health care panel that we had last week as well, where they were all talking about how our technology has made it easier for them to give patients access to care, partially because they’re not waiting on hold for a nurse that has a million other things to do than answer the front office phone, or somebody who’s running late for an appointment and can tell an agent that as they’re driving, you know, that they’re in traffic and they might be late.
So they can slot in somebody else fifteen minutes earlier. That is super interesting.
And it also yielded that session yielded my favorite quote of the entire event, which came from Nicole Rice from Doc’s Dermatology. And she said something I think none of us were expecting.
Everybody in the audience was like, what? She said, we teach our staff to sound as good as our AI. It’s like my mind blowing quote, and it’s such an such a great endorsement of what we do, but also of the technology.
What do you think? Is that something that you thought you’d hear, I guess, as you were joining from from DeepL?

Christopher Osborne
23:28 – 27:12
Yes. So I I well, first, you know, it it is wonderful to hear it because, you absolutely want that to be validated by the end consumer and, of course, our customers.
So, of course, wonderful to hear it. We are seeing things progress so fast.
And, you know, I’ve been in from the beginning, in this space, really, actually, always around this space for twenty years now. You know, we had the thing we used to call big data when suddenly we had these huge Internet sized datasets, but you could then process them, like, very fast.
We had machine learning and breakthroughs there and some scaled use cases. I was working with neural networks, and then suddenly the LLM breakthrough happened.
And I have been in the space, and yet I was astonished by how fast things progressed in the last two point five years. It’s it’s I think the best story I heard actually was that the two cofounders of Google, had founded the company as an AI company, and of was super into, like, what they could build with AI, etcetera.
They built a search engine. They learned how to scrape the web.
They built in a fairly simple algorithm based on clicks and then ranking from it, and then gave up on the AI dream because they had a super interesting business and built some of the best technology in the world. They founded an AI company, and they actually gave up on AI and said, yeah, we just don’t think this is gonna happen until about the mid two thousand tens when, you know, you started to see, maybe 2015, 2016, some machine learning breakthroughs.
But it was they they gave up on it. And then suddenly, a few things moved, and then things really happened.
And then even then, this last two and a half years has just been so surprising how quickly we’ve moved. So, yes, I did expect to see actually people say, you know what? The AI can perform better than humans.
So I did expect to see that. But it it’s wonderful to have it endorsed.
And, yes, the very interesting thing is I can look twelve months to two years into the future, and I can see, oh, actually, some of the things which were groundbreaking for us two and a half years ago when I wasn’t here, but I could see that you’ve adopted them two and a half years ago, is already becoming last generation tech. And as part of my job, my job is to balance the investment of building scalable and trustworthy secure systems for now, but then also investing in what will actually replace this generation.
And I think I think within twelve months, the agentic capabilities that we work with now will have sufficiently advanced where nobody’s really talking about generative AI anymore. And and it’s the agents will have reached maybe they haven’t reached the 800,000,000 adoption mark, but there’ll be a curve, and it’s rapidly advancing on how people are using these things in everyday life.

Michelle Schroeder
27:12 – 28:35
Yeah. I mean, I think we’re rapidly encroaching on that 800,000,000 number just thinking about our annual the amount of calls that we take every single year.
So I I agree. I think this is gonna happen.
I think we’re gonna be, surprised by the by the rate of change, in the underlying technology. But the thing that I think really cleared the path for us, and just made adoption so much easier.
And like you talk about, you know, removing friction, in in, you know, like in in the experience, like, we saw a lot less friction in the sales process, partially just because of that consumer embrace that we were talking about earlier, the consumer embrace of technology and agentic technology. We could probably talk about anything for a lot a lot longer than we did today, but it was just so great having you on.
We will definitely do more in the future. Thanks for being with us today, for sharing some of your predictions.
May they be Steve Jobsian in their accuracy. And then for anybody that does want to check out last week’s VOX, there is a, on demand recording.
You just head over to our site, PolyAI. Register, and you can watch it right then and there.
And please, subscribe if you haven’t already to our podcast, and we appreciate you being with us today.