Why Would a Small AI Company Go Public? with Lyle Pratt

Why would a small AI company go public when everyone says to wait? Let's find out with today's amazing special guest, Lyle Pratt. Now, Lyle, I'm really excited to have you here because you made a really interesting move to me. That is like when everyone is chasing VC funding and there's a lot of like this VC shell game, you went straight to IPO instead of raising private money.

Can you talk a little bit about that decision for a new AI company and why you chose to go that route? Yeah. well, first of all, you know, everyone on my team, my co-founders, my executive team, you know, we've all built and scaled and sold companies before. So, you know, we've all had the experience of raising VC money.

We did actually raise you know, some VC money, but we were very picky about our investors to make sure they were aligned, you know, with what we wanted to do long term. I think that there's a there's a variety of incentive issues with the current state of VC. I mean, if you just think about it like an asset under management game, an AUM game.

You know, let's say you have a a large portfolio company and they're worth billions of dollars and you know, you're making two and twenty on the valuation. of that company, if you're a big VC fund. So what happens when you IPO when you're when you're that large already? Well A, you get marked to market, for the first time, which can be scary. B, the VCs, lose their two and twenty carry on you know, a huge valuation piece of their portfolio.

So you know there's a there's a lot of fear mongering, I think, from VCs about you can't go public because it's too complicated to, stay public with all the rules, or it's gonna, eat up all of your time and you're not gonna be able to innovate. things like that. But I think that that is more and more becoming, just a view of the past. And it's more and more VCs talking their own book.

And you know, with AI, so many things can be automated that previously couldn't. so it's actually easier. know than ever, you know, to be and stay public. so you know, I think that's a big piece of it. you know, we'd done the VC thing before and wanted to do things a little bit differently. I think that we will be the first of many companies going public at a at a earlier, smaller stage.

And you know, honestly, I think just from a social perspective, it's the right thing to do because currently the vast majority of public investors are locked out in participating, you know, in some of the most innovative and fastest growth companies you know, in the world because they are not a qualified investor, you know, and and they can't even invest in DC rounds. So, you know, part of it is that as well. It's sort of a principled approach.

You know, we want everyone to be able to participate in this agentic boom because it's gonna be huge and people need the opportunity to invest in things like that.

I guess thirdly, it's about you know talent acquisition. you know, we've all been in the shoes of, you know, somebody who's working for a startup, you know, they have startup equity, you know, who knows what it's actually worth, you know, whether that is, you know, a small snapshot in time or in the future. and it's just much easier to actually, you know, acquire and attract talent when they know exactly what the equity they're being granted, you know, is worth.

So I would say it's those three, you know, pegs of the stool. Yeah, it's really an interesting strategy to me because it feels like when you have VC money in the larger firms you're dealing with, like they take start to take more seats on the boards of directors, they want to have a larger role in the decision making.

When you're kind of public, does that because there's so many different people investing instead of one person with a lot of control, does that mean you as the CEO and the main founder actually maintain more control? than going the VC funding route for years and years and years. Yeah, I think that's definitely a part of it. we are beholden to our investors and obviously, we want to do everything we can to, increase shareholder value, for sure.

But yeah, we we do have more long-term control over the company and the direction it's going. you know, we have more control in some ways over like, you know, who joins our board, you know, versus you automatically get a seat or something because you led the previous round. so you know, it's it's definitely a factor. I think that you know, we wanted to be able to sail our own ship you know, without needing to sort of you know change our direction every time.

Time we raised a a new round. also I think once you're public, access to capital is actually more readily available. you know, so you know, some for some reason there's this weird view where you know, once go public like you're you're done, like you know, you can't ever raise money again or or something like that, which is, you know, really far from the truth. you know, so there's definitely advantages. I mean, I'm not gonna say that there's not some disadvantages as well.

I just think that, you know, the calculus of weighing those those two things. Yeah. I think what's really interesting to me is that usually you see, especially with AI companies, they're trying to wait till the maximum value to IPO.

And of course then immediately you see this fall off or they want to IPO for a trillion dollars or like the big we always have to be the biggest IPO in history and it's always like maximum risk for the like regular investors who are coming in way late to the game after like it's been 10 rounds or 50 rounds or 100 rounds of funding and it's like You're coming in to a very late in the game.

And maybe there is a chance that it will keep going up. And sometimes it does. I remember the Google IPO and I told all my friends, you shouldn't jump on that. It's not gonna go up. Boy, was I wrong. Right. So it can be a great opportunity. but that's very interesting to kind of you let people in earlier into the game. And have you found that it changes how customers perceive you when you're trying to form partnerships and alliances.

I know you do a lot of partnership selling. Does that kind of change the equation? Because it kind of there's no mystery about the company value or what you're doing because you have the filings and all that stuff. It kind of creates like a there's no doubt area. Does Yeah. it create something there as well?

Yeah, I absolutely. it's definitely an advantage when talking to customers. transparency typically is. you know, they can see that we have plenty of cash and that, you know, they can go read R S one. you know, they know we file our, you know, quarterly, you know, to update the business. So, you know, anytime you're evaluating long term partners, you know, the more transparent you can get, you know, the better. Because that's going to make everyone participating in the partnership, feel more certain, about the direction.

So I think that's definitely, a plus. Like you mentioned, Google, when they IPO'd, well, they IPO'd way earlier than what is typical even today. it was a lot, but it wasn't anywhere close to, for example, SpaceX, IPOing at over $2 trillion, right? And so, I like to say, okay, if I'm a retail investor and I want to produce Participate in the AI boom, the agentic boom, you know, that's happening right in front of us. You know, what am I going to buy in public markets?

You know, I could buy Google, you know, they're a you know model provider, an AI infrastructure provider. but you know, they're already, multiple trillions of dollars. You know, how much are they going to grow? You know, are they going to double? You know, over what time? maybe they triple over what time? You know, maybe you could buy.

Data center plays, also, extremely capital intensive, highly valued. maybe you're buying chip stocks, it's the same game, everything is already so highly valued. what do you buy to actually get some crazy growth, over a relatively short amount of time? And there's really just not a lot of options out there in the market.

So, Vita is focused on being. an agentic AI employee play in the public markets because that's what we're laser focused on. it's not super highly capital intensive intensive, so we can be very efficient, with our organizational structure, and we can participate in this clearly, massive opportunity of AI agents, AI employees getting deployed inside of enterprises and businesses of all sizes. So I want to get into what you do in your growth.

And I think that one of the big challenges in the AI market is that everyone's message sounds the same, even though they have completely different products. You go to Zapier's homepage. We're an agentic AI native company. It's like you're definitely not, for sure, because your website said something different six months ago, right?

But someone visiting for the first time doesn't know because everyone says they're AI native, everyone says they're AI agents, everyone says either we train agent to be just like you, a digital clone or an AI second brain, or the perfect AI employee. And so because the messaging has been the same and it's the same across the board. How do you establish your USP or differentiate what you actually do and say, no, we really do it when everyone else says they do the same thing?

Yeah, well it's hard, you know, for sure when everyone is saying similar things. I think, you know, the proof comes into, you know, satisfying customer needs and requirements. You know, what we found is that customers want, you know, an expert. To help them deploy an AI employee that solves a specific set of tasks, or outcomes, in their organization. So we're very focused on providing the infrastructure and the management interface and the underlying features like outcome-based billing in order to make that reality.

So clearly, you know, many enterprise customers, small business customers, medium business customers, they could go sign up for something like you know, Grokbot and you know attempt to fulfill their need themselves. unfortunately, you know, we're still pretty far from that, especially for complex, you know, agent use cases.

And so what we find is businesses really want an expert to come in and look at their process end to end and say, okay, you know, now we can give you an AI employee that actually does this work accurately and efficiently, you know, week in, week out. And if you have problems with it, you know, you can come back to us and we will make sure that, you know, it continues to to work. it's supposed to. Also, I think it's you know clearly about platform capability.

You know, we started in the voice AI space but knew that It wasn't always going to be, we weren't always going to be voice AI focused because our thought from the beginning is that, enterprises, businesses, and partners are going to be deploying hundreds or thousands of agents, to either solve internal needs or needs for their customers. And, you know, what do you need in in in a platform to actually be able to do that?

You know, need The ability to have, you deploy multiple agents under an org. You need multi-org or multi-tenant capabilities. You need the ability to manage and see exactly what your agent workforce is doing, whether it's across your own internal use case or across hundreds or thousands of customers. You need observability, you need compliance, you know, you need security. So a lot of the same things that you would imagine that you need when you're managing hundreds or thousands of employees, like in an IT infrastructure.

A style situation, you need those same tools when you're managing hundreds or thousands of agent agents. So, you know, we started with this voice AI use case, then we rolled out our computer agents, and that's one of the things that's setting us apart quite a bit these days, is that we can give you a first world-class voice agent, but that agent also has its own computer.

It can do work after a call ends. it can manage a business workflow from start to finish, without needing to use two different systems or two different platforms. So, you know, it's definitely a fast moving industry where capabilities are changing, you know, at record speeds. you know, I think model capabilities are doubling. more than twice a year at this point.

So part of our job is to make sure that our platform exposes the best technology, you know, the best models, the best infrastructure, so that when you come to us with an agentic need, we can give you the best tool for the job.

You brought up something I think is really important, which is the idea of security and consistency. the beauty of AI is that it can be creative, and that's the danger, is that all you need is for it to say one. weird thing on one phone call and suddenly you've made a promise that now the courts are saying you have to keep up. So an AI, even though you say the AI said it, they say, well, the AI works for you.

So it still counts as an employee. And we saw that we starting to see that come into the some lawsuits. And I think that's important when we think about what we deploy, which is that you can save money up front by building your own thing. And I build a lot of my own things and I'm like I know how janky they can be. there's when you're daisy chaining, all it takes is for one thing to go down.

And the AI goes, I made my own decision here. I decided and it we'll go rogue all the time. So just that limitation and the consistency, I think that's what's really interesting. I think that's the next frontier, which is not AI can do everything, but how do we keep it from doing the things I don't want it to do?

Like how do I keep it from making up a false claim and saying we don't do refunds when we do, we're saying we do do refunds when we don't. both really big problems. And I think that until it happens to you, you don't realize that cutting corners now. And I think that's what a lot of companies are describing when they fire all their employees and they try to bring them back when they're out of it. The AI is not quite ready there yet.

There's a lot of confusion about the capabilities. And when you're deciding how to enter the market like this, especially because you started with voice AI first. And you've done a lot of work with partners. What's that process like when you're deciding our instead of going direct to customer, we're gonna go through partners platforms, do a lot of relationship building. Does it feel like you don't own the customers as much because there's a layer in between? Or does it kind of provide that layer infrastructure?

Like what does that feel like? Yeah, so you hit on a couple of things there. may maybe I'll talk about the being able to trust, what these AIs are doing. and Ultimately, it's different for every customer in every deployment, right? Because in the same way that you would give certain employees, certain access to some things and other employees certain access to other things. And you might, hold an employee at this level, you know, accountable differently than you would, you know, an employee at this level.

And agents are no different, right? So you have to first have the tools to put up the guardrails, and then you have to enforce those guardrails, and then you have to monitor. Those guardrails. And you know, that's something that you sort of get for free, you know, with our platform. It's something we're very focused on. You know, moving over to partners, you know, we think about this as the long-term game here.

You know, what we're trying to do through partners is secure durable distribution mechanisms so that we can deliver a gentic capability, agentic employees. via our partners, over the long term. So when we go and find a partner or a potential partner finds us, we're thinking about okay, what's the value of this partner as a distribution channel over the long term, you know, to Vita? You know, it's not that we're like 100% and only work with partners.

You know, we do have direct and sell direct as well. So Partners are more like a durable distribution mechanism that we can take advantage of over the long term, in a mutually beneficial way, between ourselves and our and our partners.

So, this is a this is a long game. we're only in the first innings, of of what's going to happen, with agents in the business context. we've only had harnesses that could control, you know, browsers and computers really for less than a year. year, at this point, at least effectively. So we're still so, so early.

And, you know, we want to do everything that we can to secure distribution over the long game because ultimately distribution is going to be one of the most durable modes that you can have in business. Because, you know, as software and implementing features and things like that gets cheaper and cheaper, you know, have to build your business on other durable modes. Yeah, I think that it's so interesting. the motes have really changed over the past few years.

Is like certain things used to take time and now smaller companies suddenly have an advantage in agility. And like one of the biggest motes I'm discovering is like person-to-person communication because so many people's ability to talk to each other's kind of decreased over the past 10 years. I know a lot of people they won't order pizza on the phone because they think they're going to get judged. I'm like, I don't think that's My friends used to answer the phones there.

I worked at Toys R Us, so I had a different first job, but I was like, I was never judging anyone. I was just trying to get to the end of the day. When you're on minimum wage, you don't have time to judge people. But we've lost a lot of these interpersonal communication skills. And so it's kind of becoming this rare ability to communicate or to network. Cause like we've just drifted away from it. Where it's like, trust is becoming a really valuable currency as in.

And I think that having a partnership kind of brings in that trust because you go, I don't have to audit them because someone else already has. So it gives you a really cool advantage. But one of the challenges with partners is how do you get them to want to deploy you when they probably have tons and tons of other partners in their networks?

And how do you become the one that they give attention to instead of like the one who's just at the bottom of the listing, like at the bottom of their partner's list. Yeah, well, you know, that's one of the reasons why, you know, we're getting honestly more and more selective you know, of who we bring on as partners. What we want to see is that A, you know, we're going to be the AI platform that a partner deploys, you know, AI solutions on top of.

So, you know, we definitely don't want to be in a stack with a whole bunch of other, tools that they might choose. so, a great example is our partnership with Allo Fiber, and you would think, okay, a fiber company like Allo, what do they need, agents for? Well, they're looking at their competitive space and realizing that, okay, we offer business voice services. We need to start offering agentic voice solutions alongside those services.

We have a giant MSP arm, we need to be able to use our MSP arm to deploy AI employees to our small and medium business customers. So ultimately these partners are also making a competitive choice. They know that they have to stay up with the times and be a leader, you know, in in the products and services that that the They're deploying to their existing customer, trust trusted customer base.

And so, they're looking for a long-term partnership and a platform that will give them the flexibility that they need to deploy, you know, whatever agentic solution they might be able to sell in the future, not just the obvious stuff like you know, deploying AI voice alongside UCAS seats, you know, for example.

So, you know, we're definitely looking for partners that want to partner with us directly to attack this agentic market. and not necessarily looking for partners where we're going to be, another catalog item, if that makes sense. Yeah, I think that's something that I really like about your approach, which is that some people think, if I could just get in the biggest store and it's like list anything on Amazon and see what happens. the biggest store is not always the best thing.

You want the store that's excited to work with you. There's a lot to that. So when you're working with a smaller partner or a partner who's you're the only person in your category, then they might be smaller, but at least they're going to give you that attention. They're going to put you on the end cap at the store. They're going to give you attention.

I think that's What's really interesting, and it's really very, very interesting going after fiber and these communication because the entire way we approach phones has changed so much in my lifetime. When I was a kid, I remember how exciting it was when we got a longer cord so you could walk further away from the phone when you're on the phone. It's like my kids are Well, it's a phone cord. I'm like, Well, let me tell you about pay phones or let me tell you about pagers.

And there used to be a car phone, used to be a phone that was only in your car, and used to have then you had two cell phones. So like The way we use telephones and has changed so much and the way we engage with I haven't had cable television in 20 years. what am I gonna do? But of course I have internet or I'll have two internet providers. So a lot of the ways we're dealing with these staples has changed a lot.

How do you stay forward facing when we don't really know what the next future of kind of phone communication is? Cause it's changing even separate from AI, the way we use phones changes, Every few years. my kids can't believe my computer have three monitors here. They can't believe they're not touch screens. that is they've never seen it on touch screen.

I had to learn I was watching one of them use a phone and there was they could t double tap to make YouTube skip forward 15 seconds. I was like, you could do that? Like I guess the features around for 20 years. I learned it from my kids like six months ago. So the way the younger generation is approaching technology, how do you think about future proofing or staying aware of like Both AI changes and phone changes. Yeah.

Well, it's not easy for sure, but it's one of the core value propositions of working with Vita. It's why, we have support for basically every model provider out there, right? We don't think that customers that are working with us or partners want to be locked into a single intelligence vendor, especially when, the the winner today is not the winner two weeks from now. much less two years from now. also things like you know, our voice AI stack.

So we support pretty much every AI voice vendor, every AI speech-to-speech vendor. You know, if it's out there, you can use it on our platform, you know, to make an agent that solves your use case. And our partners and our customers are trusting us to help guide them, you know, toward the best solution for their job, the most cost effective solution, you know, for their job. In the communication sector, I've been in telecom, you know, communication space for quite a long time.

So I've seen, you know, a lot of these changes firsthand. But I think if you just you know look at first principles, like, you know, how many IVRs are there out there? You know, an IVR is like a press one for this, press two for that. there's at least 400 million of them, out there, and they're all gonna turn into an AI agent, eventually. So there's also still this enormous amount of low-hanging fruit out there, in the telecommunications industry.

But I think more broadly, if we want to zoom out a little bit and think about, something that's going to continue to happen from a macro perspective, it's this idea that more and more. Companies like ours are going to look more like service companies of the past. and if we think about it in terms of dollars spent, globally for every dollar spent on software, there's six to seven dollars spent on services, and so you know, what does that look like from a telecommunications perspective?

Well, you know, we could look at the BPO industry, the business process outsourcing industry, which encompasses call centers, you know, and things like that, right? Well, in the United States, a average price for you know a call center handled minute is about a dollar eighty per minute, right? that you know AI is absolutely going to change the economics of that. And and it looks more like software providing a service. And so, you know, that's what our customers are using us for.

You know, they want some software, some agency software. to give them and fulfill a service that they expect to be done day in, day out, week in, we out, week out, which would otherwise, perhaps have been done through a combination, of software and services from other vendors.

So this whole idea of services becoming, able to be deployed by software for the first time, is going to change a lot of things, not only communications, not only how software is consumed, but Pretty much everything in the economy because it's it's just such a huge transition. previously, if you were working on a service company, you couldn't raise VC dollars because the only way to scale your service company was to hire more people to perform those services.

But now, you know, we have AI accounting firms, you know, raising money at a billion dollar valuation, you know, which would have been unthinkable even a couple of years ago. Yeah, the way we think about software now is changed. we're seeing traditional SaaS companies, even if the revenue goes up, the valuations are going down because people are just convinced the SaaS pocalypse is around the corner, which I guess they've been saying for about two years now.

And I love that you brought up the fact that the models are constantly changing. The biggest fear people have is picking the wrong one because it's feels like you're locked in, especially if you're doing an enterprise contract. And then the best one does change all the time. It's usually multiple times per day a new better model comes out. And then I was looking at the top rankings today and they're all fraction of a percent apart. It's like, this is eighty-one point zero one.

This is eighty one point zero two. I'm like, no one can tell the difference. Really, in the top ten, especially like between two and number ten, the difference you you unless someone tells you, it's like the old I always think of the old Coke Pepsi experiment, right? No one knows unless you tell people they don't know which one's which. I don't know how true that is, but I remember those commercials.

And I think the same thing, which is like the models change constantly, and there's this fear of like I lock into something, but I don't know if it's gonna be the right one in six months. It's because the market's moving so fast, it's really hard to make a long-term decision because by the time your contract ends, and we're seeing even things happening with like RAM purchasing a lot of these memory companies is like, why don't you lock into the price now?

Cause the price might go up. And it's like, Do you want to lock me in because the price might go down? What's happening here? It's like it's a lot of like this really strange forecasting because the RAM prices have spiked, but they'll go down eventually. They just have to. that's how the world works. Another company will come online or someone will make it. I look at the data centers the same way. yeah, data centers are cool, but Computers used to be huge.

A single computer used to fill a giant room and now a computer fits in my pocket. So I feel like the one thing I know is that pri technology always gets cheaper to build and computers always get smaller. I don't know a lot of things, but I know those three things have been pretty consistent throughout my life. So I like the direction you're going. I think that it really is the reason people hate like voice agents or AI agents is because they can't solve your problem.

And like people deploy these agents, they can't do anything. And I remember I used to work in a call center, and I didn't have any power. if someone needed a new mouse, which cost us eight cents. I had seven emails and like getting authorizations. And my I was really good friends with the head of support for the entire country.

And he was like, dude, I it's takes me two hours to get approved to send a mouse out to a customer who buyed a five thousand dollar computer. So like I've been on the other end of that, both sides of it. So it's like very, very interesting. And I really very interested in what you're doing because I'm More and more of my clients are working on phone projects.

They want to move away from call centers because their call centers do very little and they can't to retrain a phone agent is also a huge thing. So every time you have a change in your business, you go through huge process, whereas with an AI you can retrain in a matter of seconds. So very, very interesting.

For people that are interested in Vita, what's the best place to see what you're doing, kind of the stuff you're doing and learn more about and also maybe even see kind of what your stock is doing, which is very cool as well. Yeah, our website, vita.io, we're also active on LinkedIn and X, so you know, find us there.

You know, I would say, you know, all the things you just said about, well, you know, a a lot of things there, but you know, ultimately I guess I would finish with this, you know, people want to pay for outcomes. They don't want to pay for tokens, they don't want to have uncertainty about what their, you know, price is going to be for. some agentic workflow now or in the future.

They want to make sure that they're always using the best tool for the job and that they at the outcome that they ultimately care about is getting delivered. And so I think that's where things are going and that is our focus.

I think, you know, the companies making, you know, billions and billions of dollars selling, you know, tokens, I think that that's going to be a hard game to win over the long term because ultimately tokens are a commodity. what's not a commodity is outcomes, very specific outcomes to your business that move the needle for your business. And that's where things things are going.

You know, whether that's a voice agent handling something that, you know, was previously handled by a call center, or whether that's an agent, you know, picking up a workflow on Monday and doing, you know, all of the tasks in the queue, you know, throughout the day, ultimately businesses care about getting charged for outcomes and reaping the benefits of those outcomes. So I think that's where things are going. If you're interested in that, follow us at Vita, our ticker is B-I-D-A on the New York Stock Exchange.

So, you know, I've really enjoyed this conversation, Jonathan. Thank you so much for being here today, Lyle, for that amazing episode of the Artificial Intelligence Podcast.