Orgs establishing AI competency centers or hiring Chief AI Officers are heavy in 2025 predictions. Sit beside the fire with Damien and Nikola Mrkšić, CEO and co-founder of PolyAI, in this enlightening episode of Deep Learning with PolyAI, as we tap into the emergence of Chief AI Officers (CAIOs) in organizations, their responsibilities, and the impact of AI on business operations. Explore the parallels between AI-heavy roles and traditional executive positions, the importance of having the right vision in addition to technical knowledge, and the potential future of AI leadership within enterprise settings. Follow PolyAI on LinkedIn Watch this and other episodes of the Deep Learning pod on YouTube
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
NikolaI think their job is to make an educated guess about the place in the organization where current AI technologies can make the biggest impact to the business and making sure that the company then stays laser focused on justifying that investment.
DamienHello, and welcome back to another episode of deep learning with PolyAI. The only podcast that's mentioned by name in state bill 10 37, coming out of California, the AI safety and privacy act. Can you believe that Nicola
Nikolawould the word podcast after
Damienit? No, I just think it's deep learning. It's somewhere on page three. I don't know. I read it on Twitter. So, whatever's on Twitter is fact. No, but welcome back to another episode. I'm thrilled to chat with you. My name is Damian. I'm the host here and joining me as usual is Nikola Mrkšić, CEO and co founder of PolyAI, Nikola. Welcome. Thank you for having me again. Of course, of course. Always thrilled to see you. Today I thought we would not talk AI legislation, even though it's happening at a snail's pace here in the States and mostly worldwide waiting for somebody to set precedent.
DamienBut. I thought what we would talk about is some of the human capital that exists within organizations and who's going to be responsible for this big beast that is AI in all shapes and forms moving forward within enterprises and sort of corporate or corporate governance structures. So I've seen a lot of CAIOs thrown around in title for organizations that are building out their fundamental, like central generative or AI programs and and. It hearkens back to the era of a burgeoning focus on D and I initiatives.
DamienSo diversity, equity, and inclusion initiatives, basically modern business was way way past the finish line when it came to being able to prepare for something like that. And so what they did was, okay we're short on skills to be able to tackle this sensitive social issue. So we're going to hire an executive call them a chief inclusion officer, chief equity officer, whatever have them put together a team, have them put together a program and hopefully do something with it. So it shows that we're being good corporate citizens, right?
DamienWell, in the past couple of years, not only has this position risen, but it's fallen just as quickly with many large global brands abandoning those initiatives is the same thing going to happen with a chief AI officer. So let's first discuss what that sort of person looks like, and then talk about the longevity there and how to make sure that it stays a fundamental part of the core team. So what is a CAIO to begin with Nicola? I
Nikolamean, I don't think there is such a role, right? I think that rushing to adopt AI as if, it's a thing that you owe the company society. It's not a great or good. It's not like the AI, right? So it's just a new technology, right? And it's a technology advancing so fast. And maybe you need a prominent voice on the executive team to Steer the company towards it, right? They need buy in from everyone, right? And like the companies that will do it best or companies who's, CEO, CEOs, CFOs are motivated to, to achieve this, right?
NikolaDo they need a peer on the executive team that tasked with that specifically? Depends on the business and what products exist to make them a better business, right? If there's a transformation ongoing where AI will completely change the way they do business. Well then perhaps, right. Or maybe they need a different owner for that kind of like line of business. Right. I think the CIO is still actually the key person, right? It's someone who looks overall at the spend and the initiatives and the technology being deployed and makes, decisions that are driven by both.
NikolaThe potential benefits and the risks that weigh them all together and work to make the right decision when it comes to AI, I think we see a lot of companies at least I've heard of quite a bit of this, right? Where they think that just mentioning more of it. In earnings calls will, have a beneficial impact on their share. And, that's not actually incorrect, right? I think the longer we go, the more they will be judged on whether that AI implementation has had meaningful impact on their top line, their bottom line and other things that it can have an impact on, right.
NikolaTheir brand advertising, who knows, right. And So, yeah, I think that finding the right person anyone who's going to have the leverage, it's not that different from a transformation director, right? That role is something that many companies have, right. That aren't
Damienthose people more focused on business processes than they are technology. If you're saying that the CIO is going to serve as at least the interim CAIO on the front end, is that for buying decisions? Because obviously. Artificial intelligence is mostly found within technical infrastructures. What about ongoing management or even more like staffing towards, better business processes? Well,
Nikolayou just use the word business processes and then you said the transformation is about. Business processes. No,
Damienright. Yeah.
NikolaYeah, I
Damienguess what I'm saying is on the front end, you have the CIO is making the purchasing decisions because at the end of the day it's folding into the technology infrastructure. Right. Beyond that though, does that mean that they earn their title as CAIO ongoing as it becomes a functional part of more than just the technology. So it's going into, sales and operations, rev ops, marketing, et cetera, and moving outside of it's realm. Traditional realm, rather.
NikolaYeah. I think that at least, in our world, people buying their software it CIOs are always involved, right. But they're not the buyer. The buyer is a business line. Right. So it really matters that business line is motivated because if they're not going to change their processes and business operations fast enough to make use of it. Right, right. No, the chief AI officer. I don't know. It's a weird beast, right? Like I think that if you've got the right kind of person, you can call them, whatever. They can be a transformation director, chief transformation officer that could be the CIO themselves was visionary and driving the company in that direction.
NikolaIt could be a CFO looking to improve bottom line by, like implementing AI in certain places. It could be a chief revenue officer. Using AI to, really drive sales. Right. So I think it's it's hard to just say, Oh, now we need this. Right. To share like and reflect what you said at the start, but the I think companies were like, okay, we need to appoint this. Chief officer, right. To show that we're doing something, right. There was that moment
Damienin time,
Nikolaright? So it's a bit
Damienof an honorific, but a CAIO isn't an honorific. A CAIO is a practical leader. Theoretically. Well, do they own the budget for AI? Yeah, that's, that's a great question. Do they, so will they always officer,
Nikolaright? They're an innovation officer focused on AI. Then look, that's great. Look, we've done like a lot of business. We've been brought into many, a great complicated organization by chief innovation officer. Right. Well, or like the innovation office, right. People who work in innovation, right. Right. The job is to, give people ideas to run pilots, figure out like what could be impactful. And then, later on, there's a business owner that actually owns the budget for that, right? I think that AI is no different to other technology in that. It will be the same.
DamienI think on a cross industry basis though, that innovation office, that innovation center isn't always located. And it's more often not located within it though. Correct. Right. So, so maybe it could vary by industry, because you might find like in financial services, there's probably more innovation happening within the technological sphere. Whereas, construction or something like that, the innovation isn't necessarily happening there. And again, we're not. Oh, we being PolyAI aren't necessarily out there like crushing the construction game. Although someday we'll be there for purpose of our function, but just a broader application of AI, like it's obviously going to be industry specific.
DamienDo we find that it's more likely that whoever owns that innovation slash CAIO title, at least in the interim is going to be in it or outside of it? Are you saying it varies by industry? I think it's mostly outside of it.
NikolaOkay. I think that, you have to be technical, right? Like reasonably technical, you need to understand how it's going to fit in this technology, right? Right. But I think that this person should know enough about AI to be dangerous and know enough about the specific forms of AI being used in that company to be more than dangerous, to be like very dangerous, right? That they need to be able to tie it to the value. Right. If they fall for the values, then they will have massive impact. Then, they will, that appointment will have been a success. I think if it's just a matter of, spending AI money by, committing large sums of money to the three cloud providers and then ordering some GPUs that your teams will hopefully use, that's D minus answer to the exam question.
NikolaRight. If you can motivate some of those like business line owners, right. To implement things quickly. And really I would find that role, whatever we call it. I think their job is to make an educated guess about the place in the organization where current AI technologies can make the biggest impact to the business and making sure that the company then stays laser focused on justifying that investment. Cause in the early stage of the hype, yeah, you'll get some Money to play around with, right.
NikolaJust in the heat of the moment, 12, 18 months later, people come back asking. So what was the ROI?
DamienRight.
NikolaAnd I think really the role of that, CEO, whatever their title may be, right, is to keep the company motivated. By demonstrating value continuously, right? By showing that, well, look, Hey, we did that thing that paid off. Let's do more. Let's do more to this. Then we need this built so that AI can, I don't know, improve our pricing because by feeding on all the signals, okay. We need those signals in the right format. Please go do that for me and do it by then. Right. It's just good project management.
NikolaRight. But I think the visionary aspect here is If you're a very large company and you can invest in everything from, AI that revamps your project plans to AI that picks up your customer service goals to, like computer vision that. Integrates into your smartphone app and so on and so forth.
DamienYou
Nikolagot to figure out like what's actually
Damiengoing to work, right. And is that a chief executive though, or is that an implementation consultant that arms the business function heads that are responsible for measuring that in the longterm in the 12 and 18 month rollout, and then come back and report to. Their own division leaders, their own big companies
Nikolaare complicated, right? I don't think that's true. We have like canonical roles that, that operate this, for us it's a CEO often. Right. If you really look at the ultimate reporting for others, it's not right. Some other people sell to CMOs, right? Like we sometimes do. Right. And the fact that can happen tells you that it can be quite unorthodox. Right. And it really depends on where people can implement things that drive value. Right. I sound like a boring textbook on entrepreneurship, but it's true. Right. It really is. Cause and again, it goes back to, I think the budget, do they own the budget?
NikolaWho gets paid for it? Right. Who gets rewarded? Who's a hero. Right. And I
Damienthink that And maybe on the other side, like who gets, who gets the ax, if it doesn't work, if it fails, because it's so new this is a transformational innovation that is so transformational that if it goes down a dark road, they need someone maybe to catch the ax.
NikolaYeah. But that's such a big company. Thank you. Which means you're right, but no, you're absolutely right. But, like thinking of it as a hedge in advance is sure far away for it to not be a success. Yeah,
Damienthat's true.
NikolaSo I think it's really just a matter of if someone's being appointed, I would think of it. Yeah, sure. Maybe you can put them on the executive committee. Right. But they are someone who is really. If they're being brought in, they're not going to be someone with feuds and lands they own inside that company. Right. They will be brought in as a trusted advisor to others. Right. And. Others will have to make decisions. The CFO will have to improve, approve the spend, right? The CIO will have to get behind whatever decisions are involved, right?
NikolaAnd many others, right? Depending on the kind of products you're building, right? If you're putting a voice assistant in place, you're really going to have to talk to the customer experience officer. You're going to have to maybe speak with like brand, you're definitely going to have legal compliance, et cetera. So I would think of this as a really elevated position to be a really efficient project manager with probably a technical bend to your profile where you really. Understand what the industry you're being deployed in, what tends to work and what tends not to work out, right?
NikolaHow do those companies work? What kind of things does the body accept? What does it reject?
DamienYeah, this seems like it's an 85 15 though. More of the project manager skill set is probably going to be applicable to this sort of a role versus the technical. Cause you just need to know how to talk. Conversationally, it sounds, or at least relatively to business, goals with regards to technology, you don't need a practitioner who knows the ins and outs of machine learning.
NikolaNo, but it would be, I think, very helpful if like many executives, they come with an entourage, right?
DamienYeah.
NikolaThat's true, which actually is a good segue, let's take a good chief revenue officer, right? They have teams that have worked with them for years, right? People who do, all the kind of like subparts, people who do sales, people who do pre sales, people who do partnerships, right? And if you need them to, if they need to lean on people who know more about any sub specific thing or new developments in them, they've got people they can call. Whether they work for them or not, they are a good executive is able to stay on top of developments that pertain to their role.
NikolaWell, just with
Damienthe newness of technology. This needs to be a team, a pretty, pretty rare group, to be honest. Revenue ops has been around since the dawn of business, right? So of course, there's going to be a number of folks out there, a number of experienced resources who have this. How many experienced CAIOs are walking around there with their own rogue ragtag group of operatives, ready to step into the next corporate environment. These have got to be some real unicorns.
NikolaYeah. Look, I think that you're really looking at. At the end of the day, if, if you go back to being real, right? If a company thinks that its business model will be completely revamped with ai then, maybe they need to like, swap out CEOs and find a transformational leader from their space, from a competitor or from a smaller player who has built like a threat. I don't know. Right? Who can then come in and rewire their DNA to use. AI more effectively. Right. Again, like the analogies with cloud.
NikolaAbound, right? And there's just a lot of, if this is going to be a hundred times more impactful than cloud, whoever doesn't get on board in time, we'll have lost a few years. They'll make all the difference as their market share gets eaten up by their competitors, right? That's really what it is. Okay. That's the fear factor, right? So then the question is like appointing this person and just saying we're doing stuff with AI, if AI is not impactful and you do this, well, No harm, no foul. Right. If it is, you have to get it right. And then the question is, does the organization really buy in?
NikolaRight. And there it's not going to be one hire. It's going to be that whole exec team deciding to make this a priority. Right. And I can tell you just from all the large enterprises that we work with and talk to, there are companies that truly feel this where, you know, I can name people, right. But I can tell you that we work with extremely high ranking executives of very large companies, of course, that move mountains internally to get, things approved to find the right way to work around compliance processes, et cetera, that, stuff happens relatively quickly or happens to begin with that otherwise would just not right.
NikolaThose people are,
Damiendo you think that's, do you think that's more because we're an AI solution or you think that's more because we're an AI solution that focuses on customer experience and that's probably a more prominent. Like business executive business leader issue.
NikolaI think the fear of, autonomous systems making decisions goes past. This is true. Well, okay. Yeah. In general. Yes. Right. But also just in terms of, an automated decision for Something in production line or anywhere else, it has legal implication. It has a lot on the line, regardless of whether it's the consumer facing or not. You're right. We are consumer facing hands, all the more visibility of this. So people have to do it, right? Sure. Like we may be like somewhat of an extreme example of that. Yeah, definitely.
NikolaBut, we also do a lot ourselves to help companies navigate that. I think that one way in which we trump all of our competitors is that the name of the game for us is how do you get tens of millions of enterprise value out of AI in a predictable way. Right. And again, I sound like enterprise software, textbook excerpts won't be a good episode. They may not be a great episode.
DamienIt's fine. We'll just do a lot of
Nikolacitations at the bottom. No, but seriously I think that, you can behave in that way and then, you will. Empower that CA or whoever they are, right? Or you can have companies and, that's probably like for more of the building mindset. If they're going to develop serious capacity, knowledge teams, then you got to make the right calls, right? Cause if you spend, tens of millions developing something and then you have nothing to show for in the end. Other than talking about doing AI. Well, what have you done? Nothing. You're probably cover it up.
NikolaYou're a big company, but you won't get what you were after to begin with, right?
DamienYeah, no, that makes sense. Let's go back to talking about like the innovation center that you were speaking of and how that would translate maybe if this goes beyond a central figurehead and federated AI management model, we often talk about. When we're doing deployments within enterprises where this is truly transformation, which is a lot of them that they may want to consider establishing an AI competency center. So sometimes it's used in reference to maybe those tier one agents that find their workload to be a little less taxing and so need some additional responsibilities.
DamienRight. And it becomes more about managing the system itself in its own departmental function. So is this. Is this a compliment to the CAIO? This AI competency center, is it an alternate path or is it something that should probably stand alone? It's just a placeholder name. It's just a placeholder
Nikolaname, right? It's like the new roles emerging in your company. To use AI to do, I don't know, like to go back to my favorite simplified analogy of all this, right? Early 20th century, mid 20th century agriculture, people have tractors and stuff. They're no longer using bulls, right? Are you a chief? I don't know, tractor officer, that's the CTO I want to be. That's
Damienfor
Nikolasure. Well, who wouldn't be right. No, but but seriously, right. That person is doing the new version of the job. Right. And yeah, like initially you may want to create like a bit of a. Praetorian guard around like this person, a few more, you need to figure out a process, give them a bit of a, like safe space to work and experiment without, the full scrutiny of the whole organization. But what you really want is them outside of some innovation group as soon as possible and into the real workforce, spreading that knowledge, right.
NikolaAnd, it may be the case that the role they do for you now is new, right, completely new and different from anything you had in Yorkshire before. It's a possibility, right? I think in the context center, these are tier two agents. These are subject matter experts. They will continue to do the same when they have their work changes. Right now, they're not just like randomly sampling calls from humans and ensuring quality. They're looking at AI calls, right? They're like going back and updating knowledge bases that the AI uses to form an answer, right? They are, like doing that kind of stuff.
NikolaThey're doing more technical work and they're, they have much greater leverage of their work because now they're impacting every conversation. Right. Again, like Salesforce and sales ops is an example of that, right. Creating a new role. Yeah. And
DamienIt seems like that's a narrative that needs to be repeated over and over for organizations that are hesitant because they fear the displacement. Or not really fear, but they concern themselves. They consider the displacement potential for really, like highly effective automated strategies or technologies like, like ours in that, yes, if you are, have a bot that's going to be handling 80 percent of your calls. Effectively, then that means that you're not going to need as many agents, but it doesn't mean that those agents are out in the cold. It means that those agents get up skilled and repositioned within the organization to manage that system because it still needs that, that human touch.
DamienSo I think that's something that's really interesting and it goes back to the old thing that's circulating among, pro AI advocates out there, which is you're not going to lose your job to AI. You're going to lose your job to somebody who knows AI. Absolutely.
NikolaRight. I think we're all companies, like people are busy. Right. Right. Like you're not just going to be like, okay, that's not done and dusted. Like you it's capitalism. It's other companies will use it. They'll have better service. And then, their people will start doing something else and they'll make them a better company. Right. So they'll have time back and you should have that time back to them.
DamienDo you think. You mentioned earlier that the CAIO, the archetypical CAIO is probably someone who leans more program and project management, but then also has some technical skill sets. Obviously there's a wider swath when it comes to functional like competency centers, depending on how the AI is deployed within an organization. Do you think that these people in general? Are typically more technically savvy, more business focused savvy, and then can ramp up on the tactical knowledge that's required to manage these systems within their specific function. What do you think the skillset basis looks like for someone who's within the competency center?
NikolaI think there are shrewd executive who can tell bullshit very quickly. Okay. Fair. And I think that, like the right person for a company for this role, if appointed is someone who can cut through the noise very quickly. Okay. And that means that they probably have a level of understanding of this technology. Maybe they are themselves technical and, quick, they've got to be like quick thinking and I'm afraid of it. They must come with like at least a few trusted team members who are far more technical so that they know who to trust.
NikolaRight. Otherwise, you'll commit tens of millions to a Google or a Microsoft. You'll be training your own LLMs and you'll have your own LLM. That LLM will be like outdated in three years. So, I think that The business side is much more relevant, right? Finding allies on that exact team, people who will facilitate an environment for that AI project to be successful is I think the key thing. And that means that they're actually a really good salesperson, right? They themselves will be able to carry the team with them and to create the space for it to be successful.
NikolaCause you know, I think. Each company implements the software, like with many clients and sometimes it's very successful. Sometimes it's very slow and delayed and staggered. And it's usually down to the ability of the champion to find their way through it, right? To call in favors. To sell really hard to motivate people to reject some priorities and timelines, right? Give up.
DamienSometimes to your point, the business value is apparent almost immediately. And then sometimes it takes a little, it takes a little finesse. It takes a little silver tongue work. It takes a little sales work. And if that's the case, then that's a great point that one consistent factor is you need to be able to be a little bit of a salesman if you're really, there
Nikolajust needs to be like a prime negotiator and that they need to create the, the cover for people to do something and maybe not do something else. Because, the company as a whole needs to, reach a decision that is making a bet on this. So if three other software implementations are on hold because AI trumps that and they believe in it, well, that has to happen. And at that point, you go away and You do it and you hope it works, right? And if it works, the person is a hero. If not, you no longer have a chief AI officer. That's about it.
DamienYeah. Well, let's hope that is a title or a role that sticks and whatever its title is, let's hope that it's a role that sticks. Three to five years from now, obviously the technology is evolving very rapidly capabilities. Extend to be more autonomous. What does this role exist in three to five years? Do these roles exist in three to five years? So I don't
Nikolathink so. Like AI will be part of every product and like purchasing decisions or investments around how it's built will be down to the same people that make those decisions about technology today. Like that AI officer, whether they're like a chief officer or just the person, the AI guy or girl at a company.
DamienYeah.
NikolaIt's very likely. In places where they're successful to be the next CIO, right? Chief exec, right? It just depends on the scale and the magnitude of that. Application of AI has, and there are probably places where just won't have that much of an impact. That's okay too, right?
DamienYeah. True. Or they end up, or they just stay in procurement because Yeah. yeah, perhaps. All right, cool. No, this has been great. Really fascinating stuff. I love talking about the human elements within the enterprise and AI deployments and whatnot. So to see how that. Grows and shapes and evolves is always just an important part of the conversation, not just the technology at hand. So appreciate your time as always, Nicola. Hope we let's go back and read state bill 10 37, because somehow we found a way to tie it back to that a little bit with regards to to some of the safety stuff we were talking about with autonomous systems and see if we're actually in there, that'd be great.
DamienBut if not until next time thanks for joining us at deep learning with Paul AI, and we'll see you soon. Take care. See you there, man.
Three years ago, Simplyhealth's PolyAI deployment was only resolving simple FAQs over the phone. Today, they run voice, in-app, email, chat, and WhatsApp conversations on the platform. Nikola Mrkšić talks with Paul...
An agent that gets better every day compounds into a very different agent within months. That's the idea behind Wren, PolyAI's new release: an agent whose job is to make your other agents better. Some of what it finds...
Companies are pouring money into AI, but not everyone is getting the returns. Ian Jacobs, VP and Lead Analyst at Opus Research, thinks he knows why. In this episode, your host Nikola Mrkšić sits down with Ian to work...
When PolyAI built its Builder MCP, one client helped shape it: Vixxo Facility Solutions, one of the largest facilities providers in the US. In this episode, guest host Damian Sasso talks with Vixxo's CTO Derek Neighbors...
PolyAI just launched Dialog-RSN-1, its first audio-native model, and your host Nikola Mrkšić sat down with its builder, Matt Henderson, to unpack why it’s a game-changer for building voice agents. Most voice AI either...