Agentic AI is the future of IT

Announcer:
Ready to explore the extraordinary world of tech. Welcome to the XTech Podcast, where we connect you with the sharpest minds and leading voices in the global tech community. Join us as we cut through the complexity to give you a clear picture of the ideas, innovations and insight that are shaping our future.

Debbie Forster MBE:
Hello and welcome to XTech Podcast by Fox Agency. I’m your host, Debbie Forster MBE. I’m a tech portfolio consultant and an advocate and campaigner for diversity, inclusion and innovation in the tech industry.

So today, I am delighted to welcome to the show Dennis Perpetua, who’s the Global CTO Digital Workspace Services and Experience Officer at Kyndryl. Dennis, thank you so much for joining me today.

Dennis Perpetua:
Thanks for having me, Debbie.

Debbie Forster MBE:
Okay, so when we start, we’re really keen to get to know you just as a human, Dennis, before we understand you as a technologist. So, how did you get into tech? Was it an obvious choice from both onwards or a bit of a surprise?

Dennis Perpetua:
I don’t think it was an obvious choice for me on the onset. I had originally gone to school for psychology, so my undergraduate degree is in psychology. In the process of doing that, through coursework, I learned that my brain was a little bit more geared towards the spatial sense. And so after that I went into graphic design and then ultimately got my master’s in computer science. And so is a little bit of a mixed bag in terms of how I arrived at where I was. But ultimately, I think that combination has actually given me a unique perspective on user experience, around digital workplace services and is a great background to kind of talk about how AI is changing the employee experience within our customers and IT in general.

Debbie Forster MBE:
I think really particularly in the last few years. So it must’ve raised some eyebrows of people looking at that on the CV of the windy career. But man, since things like lockdown, it must’ve just absolutely made you look like a genius, a freaking genius to put these things together for a degree.

Dennis Perpetua:
In hindsight, it definitely justified a few things prior to the pandemic. Absolutely, it was like, so how do you go from psychology to graphic design and photography? Which was actually kind of the jumping off point, was the physics associated with the photography. I realized I was much more of a geek from a photography perspective than an artist, which then led into the computers. And I was actually lucky to be in photography at a time where digital was really cutting over. So I got a lot of the chemistry and the physics with the traditional ways of doing photography, and then I realized I was just much better at fixing the computers and connecting all of the things than actually taking the photos themselves.

Debbie Forster MBE:
Fantastic. Now you started with IBM, but are at Kyndryl. Not everybody would’ve heard of Kyndryl, they would’ve heard of IBM because I hope you don’t mind me saying that’s the grandpa of us all in terms of tech. IBM was in the space doing the space before it was a space. And then Kyndryl was a spinoff. Can you tell me what Kyndryl does? How did it get spun off? How did that work through for us?

Dennis Perpetua:
Yeah, absolutely. I think it’s a powerful heritage, if I can say so myself, right? Because there was such a depth of technology and innovation from an IBM perspective. We have that AI background, we have that software background. We have a strong services and consulting background coming from IBM. But the spinoff really allowed us to operate in a modern world much more effectively because what it did is it un-restrains Kyndryl from being able to use any technology, not just a single provider or single set of software. So we are the largest infrastructure services provider in the world. We’re about 85,000 employees and we provide services to a lot of the Fortune 50s, Fortune 100s, et cetera. And we have a great heritage in doing so.

And in the last three years, three plus years or so, we’ve been able to very rapidly accelerate to being able to work with not just IBM’s but the Microsoft’s and the Amazons and all of the other partner ecosystem that we created. So it gives us a great set of tools that we can match with our heritage and experience to provide customers really the capabilities and the services that they require to be competitive in the world today.

Debbie Forster MBE:
And that feels like the right mix because there is that really deep heritage when you talk about IBM and that deep knowledge, proper, proper geeks, doing proper, proper geeky things, but that can often lack the agility to really make the most of it that you need in today’s sector. And I think bringing that agility to it makes it really exciting.

Dennis Perpetua:
Yeah, and I like the way you say that because ultimately what it does for us is it retains some of that DNA around innovation. When you’re part of an IBM, you have that focus on creating the next thing. And now as a services company, we retain that DNA, but it actually gives us more, going back to my background, it gives us more paint in our toolbox to be able to paint the canvas with. And so, it allows us to really look more broadly across the industry to take that DNA that we had around being innovative and bringing that forward using the best solutions. And at a time like this right now, coming out of the pandemic, on the cusp of AI revolution, on the cusp of moving IT from a commodity over into a business enabler. It allows us to really actually take advantage of the heritage that we have to apply it to that broader set of toolkit that we now have access to.

Debbie Forster MBE:
Well then, I mean, let’s think about it. I know that there’s one thing you get very passionate about, it’s Agentic AI and that is such a fast moving space. And so this went from, I think when I first started doing this podcast that was edgy, that was what people talked about being at the horizon. But right now, you can’t swing a cat without somebody needing and wanting to talk about Agentic AI. But because it’s something that you know and you work deeply, it’s here to stay. It is probably becoming established, I think, across companies, across corporate organizations. But it’s a balance to be had, isn’t there? It’s understanding how do you balance that promise of the productivity? There’s a real risk of alienation and there’s been quotes out there that haven’t helped us. How do you help Kyndryl and your clients make your way through that maze?

Dennis Perpetua:
I think this is an interesting vantage point to approach this topic, right? Is how do we prevent alienation, how do we actually use it from a productivity perspective? And again, if you look at our heritage, we have a long heritage about a decade or so of AI experience. And so we’ve carried that with us just because generative AI is the topic de jour. It’s the latest breakthrough, which is a massive breakthrough over what we’ve had traditionally. A lot of the concepts still remain the same, and it’s the basic blocking and tackling that’s associated with it. But what we found as we’re looking at the current horizon with Agentic AI, it really becomes the hands and feet to make generative AI much more impactful in customer environments in the IT space. It allows us to automate things in ways that a chatbot was unable to do. It allows us to have reasoning and the ability to actually take action autonomously. And so, I think when we look at it from that perspective, when I look at it within the context of digital workplace services, we refer to it as the connected experience. How do we actually connect the employees to the overall IT experience with Agentic AI being part of that? And so we look at it more across the horizon of everything being where is AI being ubiquitous in the employee’s day-to-day job that can be viewed differently from an industry perspective, could also be how do humans interact with it? And then how do we actually enable it as a business productivity tool? How do we actually bring those things in to enable employees to be more successful?

When we approach it from that perspective, the conversation starts from a productivity perspective, from an employee experience perspective and ultimately, that connected experience. And hopefully, that allows us to get past the alienation conversation very quickly.

Debbie Forster MBE:
Good. And I think because what I’ve enjoyed looking at is you’re often talking about eliminating the friction in the employee experience, and that’s a delicate process, isn’t it? Some can be removed, some of that friction can be removed, but some can’t be removed by Agentic AI. It may still need to be improved. So if I’m as an organization, trying to think about making the improvements, how do I understand how to make my way through there?

Dennis Perpetua:
Yeah, I think that’s a good point because what it does is it makes this conversation much more pragmatic, right? So a lot of times this conversation starts out with the answer is AI or the answer is generative AI or the answer is Agentic AI. Now, what’s the question? And what happens there is we miss really being able to have a substantive conversation around the ROI and the outcomes that we’re trying to achieve. So what is the outcomes that we need to do? It doesn’t always require AI as the answer. We start typically with a service design approach. And I think that’s the most important key is what are your business objectives? What’s the service design that’s required to achieve those business objectives?

And then, how do we actually start to look at building a solution around those things? Where do the real problems lie? AI is then becoming a tool in the toolkit to help address those things. One of the stats that floats around, and it’s quoted from a handful of folks, so it exists in a number of research articles, is 80% of AI projects still fail to achieve their intended ROI. And when we look at it from that perspective, we realize that the upfront work that’s done to be able to articulate and identify what the real business problem is, is the right place to start with any AI conversation.

Debbie Forster MBE:
And I think that is the classic problem for tech is like you said, the answer is Agentic AI, the answer is the cloud. The answer is, find me the question. And I think it is that service design approach, it is that being really clear the problems that we went to. And when we look at Agentic AI and Gen AI, if we’ve not really, really clearly understood what the business problem is, what is the friction we’re trying to remove? And I think it’s interesting because a lot of what you are doing, we spend a lot of time and money thinking about removing friction from the customer experience. But I think when we’re talking about Agentic AI, there is something really, really powerful that is about looking at how do we remove the friction from the employee experience? And that requires really looking at that ROI. But also, I know that you’ve talked some about moving away from traditional SLAs into XLAs, talk to me about that because then again, it’s that measurable discernible impact.

Dennis Perpetua:
Yeah, so I think when we move from an SLA, a service level agreement over to an XLA, which is an experience level agreement, the whole hypothesis in doing so is that the employee experience does yield a customer or consumer experience outcome. And it allows us to move from IT being a commoditized set of capabilities allows us from going from a CIO that’s giving a fixed budget to be able just to keep the lights on to actually turning the CIO into a business enabler. And when we start to look at it as a business enabler to driving that employee experience, we see all of these very tangible outcomes that can be achieved. XLA is our way to manage and measure that.

And so an XLA is really based off of a handful of things. And if folks are unfamiliar with an XLA, the way I describe it as a quantitative and a qualitative set of metrics and telemetry that get pulled together into an overall score of how well IT is functioning in the enterprise. So it allows us to pull data out of systems and say, how quick is the processing time of requests of different things that are being done, and then what’s the employee sentiment associated with those things? What are the business outcomes that can be coupled with those? And that gives us an overall score that we can measure and then, drill into to be able to identify where we can actually improve the processing or the handling of that employee experience that would yield the business outcomes.

And so, I get really excited because ultimately the XLA is a way for us to have a conversation with customers around the business outcomes that they’re looking to achieve, and then to figure out what telemetry is possible to be pulled that would actually be indicative of those business outcomes being achieved. And so, it gives us a way to say – here’s all of the IT tools in our toolkit. Here’s the telemetry that we’re seeing. How do we actually apply an update to the service design that allows the employees to be much more productive?

Debbie Forster MBE:
And I think this has become crucial, and I think it has become real put on jet fuel since lockdown, because the employee experience has changed different demands, different expectations. So what you are talking about, if I’m listening to this and thinking about my company, either I am the CIO or working within that function, or I have a CIO that I might be rolling my eyes at thinking because it’s not normally a heroic, highly sought after highly sung about sort of piece, but you’re starting to picture a really different picture of the IT department of the role of CIO in the next three to five years. If I want that, if I want to start my company on that journey, where could I start? What could I do to start opening those doors to transform?

Dennis Perpetua:
I think the key and the most effective ways I’ve seen this done is by looking at it from a top down perspective and say, what are the business levers that are actually going to drive a more competitive approach to whatever it is that you’re doing? And so, I look at this across industries. I can look at it and I can apply it to manufacturing, to finance, to travel and things. So a lot of times, and I’ll use travel because it’s something that’s fairly easy to relate to, but yeah, I think we’ve all had an experience where we’ve been trying to rent a car or board an airline and there’s been some gaffe in a computer, and that leads to some impact on the actual consumer of the passengers. And so when we start to look at those actual events, in an enterprise, we could start to look at and say, how do we actually smooth those things out? How do we actually take an opportunity to eliminate them proactively from an environment? And so making sure that those things that impact the employee’s experience, when those things happen, employees get very flustered. And I could take that same analogy and I could equate it to the last time I called my bank or when I was trying to get a life insurance policy, the computer is running slow and I have the agent on the line saying, “Sorry, my computer is running slow.” So the whole idea here is how do we actually provide a proactive IT experience to the employees that are tied to achieving the business outcomes? So it does become a top-down exercise that says the employee experience is important because there’s a recognition that as we improve the employee experience, as we empower them to be much more capable, it actually yields a better consumer experience on the backside of that.

Debbie Forster MBE:
And that’s going to be, I think going back to what we were talking about earlier, this is where we can address that worry about the digital fatigue or the alienation in the workforce, because this doesn’t have to be Gen AI, Agentics coming from my job. This is about, and it’s also for the CIO, this is not how do you pull more cost out of the system? This is not about bleeding that stone dry one more time to try and save some money. This is about how are we adding impact, how are we improving experience, et cetera, which sets up a really interesting and really different dynamic, I think on both sides of that desk.

Dennis Perpetua:
Yeah, I think so and I think as we see things like Microsoft Co-pilot being introduced, the whole idea of the ROI around Co-pilot is empower employees to be much more effective at their job. When you look at the capabilities Co-pilot does, it does amazing things. I can be triple booked or quadruple booked on a meeting, and I can be relatively productive in all three or four because I’m getting summaries and I can follow up and things of that nature. So apart from stacking my calendar much more effectively, it provides a lot of other capabilities such as helping me draft emails or do some of the things that enables me to be a qualitatively better employee. And so, when we start to marry up all of the different assets or facets of AI, it gives us a greater dimension of what it is we’re able to do as employees, and it allows us to focus much more on the business outcomes that I mentioned are set from the top.

When we set those business outcomes, what those business objectives are at the top, it allows us to actually identify where the service design elements need to come into play from an IT perspective that allow us to achieve those goals. And so, it could be proactively remediating systems, it could be bringing in productivity tools, it could be upgrading various systems to make sure that they’re able to perform reliably and with the performance needed to create the digital agility that our customers need today.

Debbie Forster MBE:
Fantastic. Okay. Well that’s a lot to digest and a lot of homework to set everyone on the call to think about how they can try and change those things. And I appreciate you for sort of breaking that down for us. I love to hear from guests because you live and breathe tech all day every day, what happens on the horizon for you? So is there anything on the horizon that is most capturing your attention, good or bad at the moment?

Dennis Perpetua:
With Agentic really entering the IT conversation in earnest right now, I think the biggest element of IT at the moment is how do we actually leverage Agentic AI to automate processes further? The challenge with that ends up becoming – there’s a lot of legacy systems in the industry today, and Agentic allows us to do a couple of things, right? Most folks look at it and say these legacy systems need to be upgraded before you can take advantage of those, the new capabilities coming out. I don’t think that’s always true. I think we can create this facade or some folks like the idea of a veneer with Agentic AI to modernize some of those traditional systems.

With that in mind, to answer your question, the thing that I see Agentic is doing is it’s the hands and feet to generative AI. And it allows us to look in the arc of IT history here and say, “To me, it’s the biggest thing since the mobile phone, since the internet to social media.” And it’s probably going to be the biggest thing until we see quantum and earnest. And so this is going to be the most fundamental catalyst to changing the way IT departments behave today and the value that they can bring into their companies. So the short answer to that is I think the biggest thing right now is Agentic. There’s a lot of governance and ethics and implementation considerations that wrap around it. I equate it to the complexity of running a CHRO and a CFO department all on its own because you’re bringing in all of these digital workers that have all of the same needs of onboarding, off-boarding, performance management, ROI and costs. And as we start to update our operating procedures to be able to accommodate these innovations, the next thing that we’re going to see is quantum around the corner, and it’s really going to accelerate this thing even greater.

Debbie Forster MBE:
I really agree with you, and I think I agree with you even more deeply than I would have a year ago, because I think when we talk about what’s on the horizon, I think this looking at Agentic AI, it has become the horizon. And I don’t think it’s too dramatic to say as you did – this is one of those big game changing technologies and it does change the game in the way that the mobile phone did. The mobile phone, once it hit and people had it in their hand and started seeing it, that ripple effect, and it has that speed, doesn’t it? I can remember being given a mobile phone by my ex, that’s not why he’s my ex now, by the way. But giving that to me and I can remember the conversation saying, but yes, I’m pregnant and you need to stay in touch, but I have my pager, right? And then, roll forward three months later, it is, you will take this phone from my cold dead hands. It changed everything. Internet. And I do think, as you say, and the heritage, I think it’s very easy for the press to do this type of “AI is suddenly here”. It’s been here for a while. Okay, boys and girls, this is a very long conversation. But it is that iteration of the Agentic AI that suddenly has allowed the whole space of AI to become that game-changing of no one can ignore it, no one can pretend it’s not there, it’s how do we ripple through. And quantum, I mean, where do you lie on that? Is quantum still that holy grail that we’re all going to live and die never having seen, or do you think it’s imminent? And what is that going to be from your perspective if it does come through?

Dennis Perpetua:
I think there’s an inevitability or perhaps a need to have a much greater computing breakthrough, given the trajectory were on with traditional computing and Silicon based chips. There is a limitation that’s going to be viewed from an environmental perspective, from an energy perspective that quantum is still a promise that there’s going to be massive, massive breakthroughs on. It’s hard to kind of view Agentic as being viable in a company today based off of the ROI, when we hear that saying, please, and thank you is costing millions of dollars. And so we start to look at the rate and pace that data centers are being built out when we want to fully utilize Agentic and Generative AI, I think quantum offers a promise that is going to open up new doors.

There’s a lot of concerns there, and I’m not ignorant to the fact that energy issues are still going to exist with quantum. It takes a lot of cooling and a lot of different other things that factor in. But when you look at Moore’s Law, I apply that to the same conversation and say, there is this journey that we’re on where computing powers can continue to get stronger and stronger, and it’s going to enable technologies like Generative AI and Agentic AI to have much greater impact on the way employee experience, consumer experience, and IT operations happen in the enterprise today.

Debbie Forster MBE:
We might need a new Moore’s law if that comes through because we’re hitting…

Dennis Perpetua:
Yeah, I think so.

Debbie Forster MBE:
We’re hitting a kind of sound barrier. Aren’t we, really? And once we break through it, there’s all the other things, but at the moment, it feels like with Agentic AI, it’s going to have us bouncing off that sound barrier for a while until we can break through that.

Dennis Perpetua:
Yeah, I think so. And there’s some very pragmatic challenges with Agentic today, right? Very much prominent in the conversation is what about hallucinations? How do we actually manage change control? How do we actually deal with multi-agent frameworks? And those are some of the topics that are very real technical challenges today. There’s guardrails that can be put in place for all of those different issues. When we talk about hallucinations, we still see it. We see it actually quite prominently in Agentic AI. Hallucinations. We think of it from a chatbot perspective, and we see the occasional wonky answer. Agentic as it’s learning, as it’s updating itself, hallucinations are fairly prominent. And so, when you start to look at the pragmatic implementation details, you have to wrap those things with what we call a guard dog of sorts, where you’re able to manage those issues with additional oversight processes. And so, there’s a lot of pragmatic considerations when you go implement Agentic hallucinations, how do you manage a multi-agent framework? How do you manage the ROI? All of those things factor in. And when you say that, the sound barrier, those are some of the things that we’re up against right now. And as we start to solve for those, I think we’re going to accelerate and then we’re going to see at the same time, an increase of compute power.

Debbie Forster MBE:
And it fits into, because you described it as game-changing, and I think we’re at that point, and actually to use the analogy, I love a good analogy, when the game changes, you have to understand the rules. And I think that’s what… so the reason it is game-changing now, it is because we’re now… this is a new game. This is a new game, new rules, new ways to win, new ways to lose, quite scary ways to lose. And so, this is not for technologists to lean back and hope it will come well or that these kinks will work themselves out. It’s moving too fast to let it work itself out. We’ve got to lean in. And I like the idea of guard dogs. I like the idea of we’ve got to understand and create some rules. We’ve got to shape that.

Excellent. Okay. Now that we’ve put that bit of the world to right, we’ve put out a call to the community that we need a corollary or a new law altogether for the Moore’s Law, really productive little podcast session we’re having here today. So Dennis, I like finishing with human, Dennis. I love hearing when you are allowed to log off from the day job at Kyndryl, is there something you’re reading, listening to watching that is lighting things up for you for the rest of your brain?

Dennis Perpetua:
Yeah, absolutely. I think the one thing, and this is kind of a guiding light for my life, I’m a footballer. The European footballer, not the American guy.

Debbie Forster MBE:
The round footballer, okay? So the British and European audience go, okay, this Dennis is okay. He understands the football is around. Yes.

Dennis Perpetua:
And kicked with your feet. Yep.

Debbie Forster MBE:
Absolutely.

Dennis Perpetua:
Yeah. So I’m an AC Milan fan, so very passionate. I had been since I was about eight years old, and that was the only thing that was streaming to my house when I was on a Sunday morning. So I learned life from the game, and so, I watched football religiously. And then also, given my last name, I’m desperately trying to learn Italian in every possible avenue. And I’m mocked relentlessly about my lack of productivity there. So I tend to joke that I can programme in about a dozen languages, but I can barely speak English well, and Italian is a far stretch as a second.

Debbie Forster MBE:
So you’re working on that. So I want a similar journey, not for the football, round or pointy. I love Italy, I love the travel, et cetera. I’ll give you the AC Milan piece. There can’t be a lot of people in upstate New York who are big AC Milan fans, but okay. And I similarly, now, I mostly have a love hate relationship with Duolingo every night of, “Oh God, got to do it” – that damn green owl, that damn green owl haunts my dreams. So I suppose that I can say, and my apologies to any genuine Italian-speaking members of our audience, and I say, grazie mille signore. Good luck with that and I can’t wait to hear how you’re getting on with your languages. And Dennis, thank you so much for joining me here today. We’ve covered just about everything, and it’s been such a pleasure talking to you.

Dennis Perpetua:
Prego e grazie mille to you.

Debbie Forster MBE:
Thank you for listening. If you’re a tech innovator and would like to appear as a guest on the show, email us now at [email protected]. And finally, thank you to the team of experts at Fox Agency who make this podcast happen. I’m Debbie Forster and you’ve been listening to the XTech Podcast.

Announcer:
Keep exploring the world of tech. Subscribe to our podcast and never miss an episode. To discover more opportunities for global B2B Tech brands, visit fox.agency today.