Techstrong TV March 11, 2026

Hey, everyone. Welcome back here to Techstrong TV. You know, I haven't had my friend here on, I don't-- It's gotta be at least eight, nine months, maybe more.

Uh, you may know him, Tyler Jewell. Tyler and I know each other now, I don't know, Tyler, seven years, eight years, maybe longer. Maybe more, yeah.

Yeah, it's been a while. But Tyler's also the CEO of a company called Akka, which is a really interesting company he's gonna tell you about. But speaking of interesting, Tyler's led sort of a charmed, interesting career.

Tyler, share if you don't-- I don't want to embarrass you, but share if you don't mind a little bit about your career path here. Oh, wow. Uh, well, I spent, ten years, as product management for some really interesting dev companies, BEA, Oracle, MySQL as well.

And then I've done, four startups. Uh, this is the fourth one that I've run, and all three previous ones were also dev, dev platform related, and they've all been acquired, by some really interesting companies, along the way, and it includes WSO2 and CodeEnv, which was a company I started in the cloud IV space. And I've been an investor.

Uh, started investing, maybe in two thousand and nine, I think was the first investment, and I've done about fifteen investments and deployed, about a hundred and fifty million dollars of VC capital, and, all, all again in dev-developer related businesses. I, I think there's just something interesting about dev platforms. Very cool.

Hey, time out. Tyler, are you expecting someone in here, Jack Bar-Barlik something? Uh, no.

I mean, it's probably the PR, PR person, so he doesn't, you know... You can, you can- Should I let him in or just ignore him? Uh, you can just ignore him.

It's fine. Yeah? Yeah.

All right. Tell them you did it, I didn't. Okay.

All right. Paul, we, we took a little break, just mention. Okay.

You know, Tyler, you're, you're being humble, but, a little bit, a little modest. But yeah, you've, you've, you've-- you know, the very kind of definition of serial entrepreneur and startup guy who has been at both, you know, at t-the board seat level as the investor to the hot seat level as the CEO founder, right? Trying to make it happen, and, and you've had a tremendous amount of success.

Kudos to you. Thank you. Um, I remember what the first interview I did with you when you first joined Akka, 'cause Akka was a company I didn't know- Yeah ...

to be honest. Yeah. You know, you, you may be aware of them.

Mm-hmm. Make our audience aware of Akka. Give us the Akka backstory, Tyler.

Oh my gosh. Well, the company is now almost twenty years old. We're just, yeah, just shy of that.

Started, by our founder and CTO, Jonas Bonér. He's out of Stockholm, Sweden, of all places. And, it was originally named Typesafe back in the day, and then it became known as Lightbend, and then we rebranded to Akka.

But Akka was an open source framework that Jonas invented, to make it simple to build concurrent applications on multi-core compute. And, twenty years ago, that was a really hard programming problem. Uh, it required almost a graduate level of computer science, and he simplified it, using some nineteen seventies computer science called Actors, and the framework became wildly popular.

Uh, it's been downloaded a billion and a half times. There's over a hundred thousand systems that have been built on Akka. On any given day of the week, two billion people use an application that has Akka under the covers.

And, and what it does is it's a distributed systems programming framework that pushes your logic and data out into the application, and it allows the application to decouple itself from its underlying infrastructure. And when it does that, it inherits these properties that allow it to adapt to the unexpected. And so when there's a network disruption or a hardware failure or a burst of unexpected user activity, the system, the Akka system knows how to adapt to that.

And as a result, our clients build systems that get, no downtime. They have like six nines of availability, massively numbers of concurrent users, huge data pipes, as in like petabyte scale of data processing and synchronization. And, and it's been used.

It's, it's basically battle tested. It's been hardened. And over the last number of years, we've, been transforming the company from an open source development framework to a, platform, that we now sell to the chief AI officer, an AI platform, for, building, building and managing, complex agentic AI systems on that.

Why am I not surprised? You know, you, you, went there and, and, and got right into the heart of where we are now in the world. And, and it's certainly- Well, it was a- ...

what an exciting time. " But I said I'm not surprised. I, I- If anyone's gonna do it, Tyler's gonna do it.

But go ahead- Well- ... tell us how ... it turns-- I mean, an agentic AI system, you know, where you're gonna unleash a swarm of agents to solve problems on your behalf, that is a distributed system.

Y-you know, you've got these agents that are talking to each other over a network. That is a form of coordination, and they're sharing context. That is distributed state.

So, Jonas didn't realize this, but he has been building an AI platform for twenty years. He just didn't realize it until AI, AI came and knocked on his door. Unbelievable.

Right place, right time, man. Good for you. Thank you.

So of course, the stuff w- you know, we're, we're playing with agentics here in, in Techstrong. I'm, I'm really-- I'd tell you about it off screen, but I'm pivoting the whole company around agents doing a lot of the dreary work. Mm.

Right? The, the, the nuts and bolts of taking videos and editing and posting and descriptions and, you know, all of the stuff that goes with that, and around our webinars and our newsletters and content. You know, it's silly not to, just as you would, Tyler.

Absolutely. You guys recent- Yeah, I mean, we- Go ahead ... we, we, we ourself have been, a, a huge embrace of, AI, AI first, task work, particularly in our engineering org.

And our engineers, have really kind of encountered, sort of two, two realizations. One is, they're about four times as productive as they were a year ago with it. That, that is, y- you know, just stunning.

And most of that productivity has really happened in the last few months. It, it seems, it seems like Claude Code every three months is doubling in its competency here. It's, it's really remarkable.

I, yeah. Yeah. I agree.

Yeah. And then the sec- November, December, I think was watershed. And not just Claude Code, Codex too, I mean, to give them their fair share, right?

Claude Code I think is the developer's tool of choice, but, but OpenAI Codex around the same time also made a bit of a leap. Made a bit of a leap. Yeah, I mean, and they're, they're, they're obviously all-- everybody's learning from each other o- on this.

Oh, yeah. And that, that is somehow accelerating the flywheel effect that's going on with it. But we, we actually have this sort of second realization that, that has happened for us, which is there's a, a cognitive overload that, that m- you know, a lot of the people on our team are dealing with because we're now doing so much work, you know, within an hour that they're-- people are struggling to try to figure out, "Well, how do I just juggle doing-- being able to do this much work and staying on top of that," right?

And so, y- you know, the people who have ADD, ADHD in our group are having a ball of a time. Hello. " I just try to keep all of my sessions, you know, in, in some sort of order 'cause I, I hop, like, yeah.

I-- you just described my days, Tyler. Um, hey, you guys recently though engaged with a company called Manulife. Uh, I, you know, kind of r- we want to call it a case study, an engagement, a customer engagement.

Yeah. Um, what-- you know, let's hear about that. Well, Manulife is one of the world's largest insurance providers.

Uh, they're based out of Toronto, a Canadian company. They're, they own, John Hancock in North America. Oh, okay.

So John Hancock is, is kind of, their North American insurance operations. But, but Manulife is a global beast. Uh, they, they operate in thirty countries around the world, you know, and a quarter, a quarter of their profit comes out of Hong Kong and mainland China, so they've got a massive presence in, in Asia.

And their, chief AI officer and their, chief technology officer is leading their company through an AI transformation, and, and not just AI, on the productivity side for employees, but they are rebuilding most of their IT systems to be a- agentic powered on that. And so they needed to lay down a platform for what will be over, two thousand, developers and data scientists to have a consistent, repeatable way to build agentic systems and then, you know, get them into production safely given the very strict regulatory controls that they're under. And, they partnered up with us, Deloitte, and M- Microsoft, and, Arize, and, to ba- basically deliver this platform.

And Aka is the agent- the agentic runtime, and we're deploying Aka in over twenty regions around the world and in, in a live HA/DR configuration, so that what you can do is these agentic developers can come along. They'll use Claude Code to design, a complex agentic AI system. It will test it, and it will be able to deploy, into one of their regions, and that region will have the active, active replication to other regions so that it has full HA and DR, so that, that you can-- they'll be able to run, run these agents all around the world, and they'll never, they'll never fail, sort of thing.

I love it. I love it. 'Cause in-- this is what we're short of right now.

We're short of real-life stories of not just people tinkering. Like, I consider a lot of what I'm doing here still at the tinkering stage. Yeah.

I haven't quite, I haven't quite- I, I, I- ... industrialized it ... y- y- you know, I, I would, I would agree that a lot of, a lot of AI is still in the personal productivity space.

Uh, we, we, we are focused on enterprises and, and particularly enterprises that are in heavily regulated environments like banks, health- Mm-hmm ... healthcare industry. Um, and, and when you're in those heavily regulated environments, there are very, very tough, bars that you have to, pass on terms of AI reliability on that.

Um, and, and w- you know, it's, it's kinda funny as despite all the hype that you hear on AI, when we engage with enterprises, most enterprises are exactly in that spot, which is they know that there's something powerful here. There's-- they know there's something they need to do, but they're not quite sure what to do, and they're wait- they're waiting to see some of these enterprise stories come out the door because they're, they're like, "And we wanna see what was successful and what wasn't successful," so they can model that and mimic that. But isn't that the way it's always been, right?

That, that, the, the crossing the chasm model, right? Yeah. The thirty...

And, and let me see what my peers are doing. Let me ask you a question, though. One of the things we're running into and we're hearing a lot about is security around these agents, right?

We're- Yeah ... we, we've installed some agents, and as I, as I mentioned to you, we're, we're doing a lot of things. Now, we're part of Futurum, and the Futurum IT team-I asked, I went to them 'cause I didn't wanna do it as a skunkworks.

I said, "Hey, we're using these agents. " Yeah. They didn't have any.

" Yeah. And of course, you know AI could be verbose. Oh, yeah, yeah, it gives you, gives you a huge dump, yeah.

Um, and I'm looking at it. I'm saying, "All right, I can see-- You know, I-- Most of them make sense. " Yeah.

Some of them I think are wearing belts and suspenders at the same time, not necessary. What are you seeing, Tyler? You're, you're...

You know, in highly regulated industries, I mean, that's security's job one there, right? Yeah, yeah. Yeah, I mean- What are you seeing?

And y-you know, y-you, you frame it as security, but the way the big enterprises look at it is it's a, a risk management concern o-on that. All security is. And, and, and risk management is actually, in some cases a legal, a legal function rather than- Governance ...

security or infosec, where they're just looking at- Mm-hmm ... y-you know, kind of the, the strict isolation of it all. But, a-and it was kind of interesting.

When we, when we filled out the RFP for Manulife, because there was an RFP process, the old classic one, the, the security and risk management questionnaire was three hundred and sixty questions deep. Oh, Jesus. I mean, it was, it was, it wa- it was very serious.

And so what, the, the, the way, the way I kind of frame this for, for people is that you have two obligations if you're going to de-design an agentic system that's gonna run in your enterprise, right? You need, one, you have to have a system that, self-explains, because your, your number one regulatory requirement is that, every decision that the AI makes has to be, expl- explainable. So you have to have chain of thought reasoning.

Uh, you have to have, immutable tracing, so you have to be able to trace all the interactions that took place. There's, there's these things called interaction logs and intent-based logging you gotta do. And, and there's this sort of causal, causal analysis so that whatever decision was made, whether it was a good decision or a bad decision, you have to be able to trace back to where did it go off the rails.

Like, if it went off the rails, at what point, and what caused that? Um, and, and furthermore, as part of that, you have to capture, all the access controls that every agent had at any point in time, because agents are, not a reflection of the user. An agent is its own identity associated with that.

And so these agents have these sort of dynamic, constantly changing access rights. So on the self-explain part, there's just all sorts of things that you gotta do, and that's a regulatory requirement. Then, then you want to have a system that can self-contain itself.

And so a self-containment system, one is, you know, at an infrastructure level, it needs to be kind of a zero trust, impenetrable network on that, so that you can't have any sort of outside actors influence it. Uh, but then you, you gotta do the basics where you have guardrails and, sanitizers like PII sanitizers associated with it. But then ultimately, where it lands at the end of the day is that since you have, these sort of teams or swarms of agents that are independently making decisions, you have to put a, a, a, a, a basically a policy enforcement engine that teaches the swarm on how to contain itself, and so that you can layer down all kinds of policies.

They could be ethical policies, cost policies, behavior policies, whatever that may be. And the swarm then takes an obligation, a contract among itself, to make sure that those policies are never breached, or if they are breached, then what is the cure a-associated with that? And so, so if you, if you look at the system in those two lenses, self-explain and self-contain, then you have addressed the security problem that, that, y-you know, and you're gonna pass regulations on it all.

I love it. Tyler, I could talk to you about this stuff all day, but we're only supposed to go fifteen minutes. We're probably over time.

I apologize. That's all right. I'm in Puerto Vallarta, where we just had, you know, that little experience with the cartels, a week ago.

Oh my God, yes. So, you know- Well, thank God you're okay. I told you my first interview this morning got cut off because of an air raid, and a, a, a, you know, a, a bomb shelter, warning.

Now you're in Puerto Vallarta and- I- Jeez. I- I'm glad I'm sitting home in Boca, man. You're-- Y-you know, it was really wild.

We're having breakfast, you know, and it's-- In Puerto Vallarta, it's just a lovely place. And- Yeah, it is ... you know, we're walking back to our, our condo, and you know, there's some guys with machine guns and ma- and masks, and they're just setting these, all these vehicles on fire, hundreds of them all over the place.

Oh my God. And I have to tell you, they were the, they were the nicest cartel members I think I've ever come across- Yeah ... because they were, they were making sure to get the public out of the way.

They weren't doing anything threatening. It was, it was- You know what? You gotta-- There's honor among thieves, but- There's honor among thieves ...

hey, be careful, dude. Who, who am I gonna talk to? Come on.

Get, get home. Be careful, please. It's a, it's a beautiful place.

It's very safe here. Yeah. Oh, Puerto Vallarta is-- it's, it is beautiful.

And it's a shame. You know, I wrote an article, it's why we can't have nice things. It's, it's-- There's so much good in our world right now.

There's so many exciting things, and then you get things like wars and cartels and violence and all the rest, man. Anyway, Tyler, it's good seeing you. Keep up the great work.

Come back on and keep us posted about what you're doing, please. All right. It's been too long.

Take it easy, Alan. All right. Tyler Jewell, CEO of Vaka, here on Techstrong TV.

We'll take a break. We'll be back. Now, as you continue to move to your new infrastructure, you've got, you, you have production nodes and, all embedded on the new, the, the new operating system, the new management platform.

What are you doing differently now carrying forward? You know, of course, we've talked about digital twin, to a certain extent, but, you know, you've talked about integration with CI/CD processes and so forth. Yes.

What do, what do day two operations look like for the whole team now on the new platform and infrastructure? Well, c- currently, you know, on, on the new infrastructure we have, the, the team now it's becoming more, you know, data... The day-day two plus, it's becoming more, more routine kind of changes, as we are, of course, moving the, migrating the data centers and, and building our- Mm-hmm ...

existing data centers on IDA. But, but also from an architecture perspective, we are now trying to do more in terms of, you know, implementing more AI functionality- Mm-hmm ... to, to now, to tell us more about the operational, you know, if something goes wrong, if an incident happens, d- don't give me raw, raw information, even if it's correlated.

Even if it's- Sure ... correlated. Give me, give me an RCA.

Give me an RCA, and point me to what, what could be the problem and what could be the impacts. So may- maybe something has happened, but actually there is no real impact to the network. So a link might fail- Mm-hmm ...

but there's enough redundancy that we shouldn't worry about it. Sure. Tell, tell me I don't have to worry.

So we're trying to build this more of an intelligent intelligence that there is some entity that is... entity, that is aware of our topology. Sure.

Of our topology- Yeah ... and our network that can actually give us useful information about incidents and the results. But the main, the main thing that's interesting also in the operational aspect is I, I, I talked to the operational team and, you know, they, they, they have a huge reduction in, in incidents, you know, like, like 80% to what- Right ...

we had before. Yeah. Uh, not, not only, not o- the tool is part of the aspect, but the, the, I think the i- the, the net ops and tracking and, this idea of doing things before you implement them helps a lot in avoiding, avoiding problems.

So, so the, the operations team are saying that most... They, they, they are saying to me that most of the time they, they don't find the same problems that, what, what they found before. The fabric is very stable.

Hmm. Very consistent. Stable, consistent is as, as expected.

There is no- nothing that surprises them. Most of the things they're dealing with is maybe, access to, to servers, load balancers, firewalls, physical, physical stuff that is going on in the data center, but not from a, a design or, or, or a routing or, or this kind of, perspective. So i- it shifts, it shifts their, their, their efforts, you know, away from debugging, you know, lower level- Sure ...

details that the fabric should, should take care of. So it's a, it's a different, different type of challenges, but much, much less than before. Much less before.

So that's a great example of the tooling decisions that you make really having an impact on your operations team, right? And you're dropping 80% of your tickets, that's, um... and, and I know, you know, your leadership has, has cited the same statistic, you know, 80% reduction in tickets.

That's, that's huge. That's certainly impactful. Didn't you also make some decisions on how to, implement Ida?

Like you could have gone with a local install for Ida, but you opted to go with the SaaS option. What were, what were some of the thoughts you had to, to weigh there, and why did you ultimately choose the, you know, Ida SaaS version? Okay.

So, so the Ida SaaS really, we, we as a network and, you know, network team, we're not in the business of running Ida, you know, or, or we want to make use of Ida. Sure. We want to architect around it.

So, so we offloaded that, that effort to the Ida SaaS team, where they, they can implement it in the cloud, whatever, what- whatever cloud they want. They want to go to- Sure ... to GKE, they want to go Google or AWS, whatever they want.

Right. We don't care. We just want the connectivity to it.

We want f- you know, the upgrades to happen. We want them to monitor the health of Ida itself. You know, may- the, the Ida SaaS comes with many extra add-ons, you know, with RCA, you know, correlation, they do some kind of AI into, into giving you in, you know, analysis of anything that happens.

So, so that's, that's also a plus that comes with. We really didn't want to, to think about Ida too much. We wanted to rather use Ida, architect around Ida and, you know, think about our, our architecture more and building more enhancements to the solution itself.

So Ida SaaS in that perspective, saved us a lot of effort and, and the team there was v- very helpful for us. So we, we like that. We, we, we were, we're glad we went that way.

Not in the business of actually running Ida, you just want it to work, right? That's, that's a great- Yes. Yes ...

encapsulation of that. Very good. Yes.

Well, what else is in store in the future here? You know, you've, you've laid some really important foundations here for, you know, making real improvement for the resiliency of the network and the lives of the operators. Um, what, what's coming down the pike?

What, what are you gonna be able to do in 2026 that you couldn't do in 2025? I think, I think now we're con- concentrating on enhancing, enhancing the solution, using maybe more AI that is nativeYou know, native to IDA. IDA's coming up with lots of capabilities- Mm-hmm ...

that are AI related that, we can chat, we can chat with IDA, we can talk to it. Right. We can tell it what went wrong, what, what's happening.

So that part, you know, we used to actually, we build in our solution, but a way or around IDA to g- to, to fill in that gap. But I think in, in the next release, we're, we're gonna be merging some of our tools with IDA, and maybe merging our automation into IDA itself. IDA has the capability to write your own apps.

Sure. We would-- I, I think that's something we would like to do is, you know, bring in our, our automation and pipelining to, to, to be within IDA, to be within... " Maybe we can merge something together there.

And also more on the operations aspect, to get, to make operational RCA more intelligent, which IDA will add, and we'll see what it adds and try and fill in the gaps there. As well as, as you know, we're still... Nokia's growing.

Mm. Mm-hmm. We've got, we've got acquisitions and stuff like that, so there will be more data centers and, and as I said, our, our architecture now is very modular.

It's not flat. It's not like, doesn't look like a data center. It looks like many small mini data centers connected together.

Sure. For sure. So we're able to absorb, absorb those changes, and that's, I think, will be a main task for us next year.

Ahmad, thanks so much for talking with us today. It's been a pleasure. I look forward to hearing how things progress in 2026.

Thank you very much for having me. Um, it's very exciting for me, and I'm very passionate to, to, to talk about our work 'cause it's, it really is exciting and interesting. It comes through in every conversation I've had with you.

So thank you. Thank you for sharing that passion and turning it- Thank you ... into real concrete, you know, this is what you're actually experiencing in the new, the new network infrastructure in, Nokia Enterprise IT.

Thank you very much. Thank you for having me. Hey, everyone.

Welcome back here to Techstrong TV. You know, I'm really happy to have my next guest on. I, I gotta let you guys in on a little secret.

We actually were gonna interview Amit yesterday, and we had to abort the interview because of a air ra- air raid warning, shelter warning that Amit, received, and so we had to stop the thing. And it-- I'll be honest, it's not the first time that's happened to me is, is, you know, uh... I, I interview a lot of people in Israel, and, and when the Hezbollah and the Hamas and everything going on over the last two years, it's happened.

Interestingly, though, Amit, I-- right after the interview I couldn't do with you, I did another interview with a friend of mine who happened to be on vacation in Puerto Vallarta, Mexico. And there they had cartel members setting cars on fire and, you know, and dem-- you know, making all kinds of things. I just...

I'm glad that I'm sitting here at my desk. I don't have to deal with all this. But thank you for coming on.

It's great to have you. Um- Thank you for having me. My pleasure.

io is a company we've covered here at Techstrong for a long time, since I started. Um, you're the field CTO there, but give people a sense kind of of your journey, how you came to be the field CTO, maybe your background, skill set, and then we'll talk more about Mend and some other things going on. Uh, sure.

Uh, so hi, everyone. Uh, my name is Amit. People call me Cheetah.

I come from a background of, cybersecurity and reverse engineering. And, in my previous company, we built a bootstrap business around the container scanning for security. So we scan technology called container images to, like, find whether a vulnerability is exploitable or not.

And actually, Mend bought our company, and that's how I joined Mend. Wow. And if you don't know, Mend is an application security company that grew recently also to be an AI security company.

Uh, and since then I joined Mend, I became the field CTO, which means, I'm lucky to speak with our largest, customers and to help them, with their product and their security issues and to get more value, and also to learn from them how we can improve what we're doing here and to make better security that is tailored to their needs. Uh, and I get to talk, with you, Alan, which is, amazing as well. Thank you, and my pleasure.

And, and, you know what? So field CTO is, it's fairly new. Maybe in the last five to seven years, you're starting to see a lot of people with...

Not a lot, but you'll see that as a, as a, a title for people. It-- Because back to like when I was starting companies, right, CTOs came in two flavor. There was the VP of engineering CTO who kind of ran the development team, and then there was more of a market-focused CTO who really spoke to a lot of the customers and became sort of the, the, the antenna, the, the input from the market that can then got translated back into how this becomes product, right?

And how this becomes service. And, and out of that grew, I think that's the field CTO heritage, right? From, from that line of, of it.

io is a, is a application security company, and now, of course, an AI security comp-- everybody is AI today one way or another, right? But, you know, there's, there's AI washing, where we just put AI in front of things, and then there's really working in AI. And, and that brings us to what I wanted to talk about with you today, which is Mend reached-- recently launched a system prompt hardening feature service.

Uh, it's not a standalone product or is, is it? No. So it, it, it's part of a much larger product, which is security for any AI application.

Mm-hmm. And so I, I'll tell you a bit about what we spend a lot of our time in the recent years now in Mend. So we realized that because of AI being such a big trend, a-any, company almost build their own AI products to sell to their customers, whether it's a chatbot or AI behind the scene that does stuff.

And there are new security concerns around it, and you wanna build these AI systems to be secure, compliant, and we try to give you all the tools to make it easy for you to secure, your own AI tools that you sell to your customers, so they can trust it. And, and we just released, what we call system prompt hardening, which is a... basically any, chatbot or LLM that we use today have prompts behind the scene that instruct the AI how to behave and what to do.

And, like, it can have, like, security issues or you, you-- like for example, you don't want to expose sensitive data there because people can leak that prompt by talking to the chatbot, or you want to help the chatbot to understand, the AI to understand w-what it can and cannot do security-wise as well. So we, we help companies do it automatically because when you have tons of developers, it's hard to control what they do, and that's what Mend is about, to help secure code. Love it.

Love it. Now, when we talk about hardening, hardening the, the, the prompts and, and all of this, and, you know, bringing risk management is the best way I could describe it. Bringing risk management.

Th-the part of the problem is I saw an interesting survey. Sixty percent of workers, like knowledge workers, right, people on computers, are being given access to AI. I thought that number was low.

I bet you eighty percent of people are probably messing with AI. But of that sixty percent that have access to AI, sixty percent are not using it regularly. " You know, they maybe, you know, did a, a, an experiment, and then went about their business, right?

I... Is that number-- that number's gotta change, I assume, right? Uh, people gotta...

But how... I, I guess the question I'm asking is: How important is it to harden our prompts now? How hard-- how important is it to bring governance to this whole...

Because it, it's kinda... You, you don't wanna kill the baby, right? Or throw the baby out with the bath water.

You want to encourage the people to use it. Definitely. You know, where's the balance?

So you definitely don't want to stop innovation, and you want to help- Right ... people a-and help enterprises to adapt more. A-and actually, a, a lot of the reason that large enterprises are afraid to incorporate certain tools is because they're afraid of the security implications of it, and we see ourselves a bit as enablers.

Now, you're correct that today a lot of people still don't heavily based on AI. A-and but it doesn't matter for hackers, because hackers are going for the low-hanging fruits. Yeah.

So even if you have, like, only one percent of the organization that is vulnerable, hackers will go to that one percent, and that's how they'll breach into your organization. From there, they'll go and start stealing information. So you wanna take the, the weakest point in the organization and help protect it.

And, and today, it's starting to become a AI, especially with risks such as indirect prompt injection, which, like, I, I don't know. I think it's something that a lot of people have heard of the concept of prompt injection, where you just convince the LLM, the AI, to do something it shouldn't. But then it becomes more complex when it's indirect.

When I ask for something trivial, the AI goes on the internet and reads a website to do it, but then this website convince my chatbot to do something it shouldn't and to leak information. And this is a super complex issue that we're gonna-- we're dealing with right now, and we're gonna continue to deal with as agents do more and more stuff for us. So definitely important.

Definitely we don't wanna stop innovation a-and just be an enabler for it. Let's talk... You know most people, they think of a prompt...

Look, a lot of people talk their prompts in now. They don't even type. You know, they talk to their AI, and so they don't see it visually.

Other people type, but they just see what they type, and they see what the AI types back. What they don't realize is the hidden instructions, the silent stuff that's going on in the background when you hit Enter in your prompt. Can you talk to us a little bit about those hidden instructions, if you will?

Sure. So any chatbot that we use today comes with a few layers of text, basically prompts that instruct it how to behave. So it doesn't really have a sense of self.

But for example, when you use ChatGPT, at the beginning, there's a huge chunk of text telling the AI model, "You are ChatGPT. This is the time today. You should behave like this.

You can do X, Y, and Z. " Then they paste your text that you messaged. "And here are the previous messages.

" So there's a lot of text going into each AI model, and it's fairly interesting. All these prompts are leaked online, so you can actually Google it and find online all the prompts of different chatbots and see what kind of information they see about you. And th-that's what make certain models behave differently than others.

Also, they have, like, different levels of intelligence, but also the prompts, right? Uh, and not only that, there's also smaller prompts hidden. For example, we let today a lot of agents to search the Google for us, search the web for it.

So they get a tool, that they can execute a search with certain keywords, and these tools come with instructions how to use it. It's more text that goes into the model. Now, the more no texts we have there, there's the risk that the model won't behave like we want it to behave.

Uh, and if, especially as we move from just question answering to doing tasks, right? So if I, I give my agent ability to delete files, to send an email, to, I don't know, delete my database or change data in production, these dangers become more and more like, complicated. Maybe it reads a tweet online that convince it to delete my, organization data by-- database.

So, that's, that's a bit the, like the background for prompts and why the fact that the agents just get text as an input is complicating things. Yeah. Agreed.

Agreed. Um, look, I'm not gonna ask you to say if so which prompts are more secure than others 'cause we'll wind up, you know, we're not, we're not gonna say Claude's better than Gemini or Gemini's better than OpenAI or what have you. I'm sure they all have their unique weaknesses and strengths.

They're all trying. " They all try. But, you know, in the, in the, in the race for progress, in the race to go forward, in the race to get users, unfortunately, security isn't always top of mind with them, and that's why you need this type of, of service here from Mend to, to help with this.

Um, you know, I mean, got a little time. So we're going through our own thing here in Textron. I think I bought three Mac Minis this week for the team, right?

Amazing. That's-- 'cause that's, that's the new thing right now. So now we're running AI locally.

We're running AI on the edge, if you will. And who figured? Everyone said Apple missed, Apple missed the AI.

They didn't have a chem-- a chat. They didn't have a LLM. Well, turns out they're the hardware of choice for, for AI maybe, right?

But, you know, so we're putting them on standalone machines. We're trying to, you know, as much as we can isolate them. Just so we start with almost a zero trust, right?

And then turn on certain things as needed. Google is one of the things 'cause we gave each agent a, an email address, if you will, and, you know, access to Google. Um, it-- and we, we're-- and the funny thing is we asked- It's a huge, it's a huge security issue, by the way.

Oh, I know. So, yeah. You know, what-- so what we're doing, I don't wanna tip too much online here, but we actually-- each agent runs on a dedicated Mini.

It runs in a VM, not on the Mini itself. We, we install it in a VM on top of that, like another abstract. We have a zero trust that we give it nothing, right?

And then we slowly build up what it can do, and, you know, where it can load and so forth. But I will tell you, a-and I, I've got thirty years in security, right? I'm not just some guy who does interviews.

I, I started security companies. The-- it's hard for me because what it can do is, is great. I mean, we're, we're using it to starting automating video editing, posting the videos, you know, vi-- editing our articles, newsletters, all of these things.

But, but the threat security-wise is real too. And so I don't know if we... Uh, you know, and how are we finding out what's best security practices?

We ask the AI. what's the best thing to do here. So it's like, I don't know if I, you know, I, I'm trusting the AI to tell me how to secure the AI.

It doesn't make sense. What do you, what do you-- what's your advice? So I, I like to quote, OpenAI's chief information security officer.

And we don't really have, hermetic solution, so you, you have t-to have multilayer approach to that. Uh, and it, it got... It, it, it ranges from if you use OpenAI, they train their models to be less susceptible to prompts, sp- malicious prompts, and hijack.

You add tools like Guardrails that try to listen to all the conversation, all the things that the AI is doing, and drop messages that are suspicious or, like, e-executions of tools that are suspicious. You have prompt hardening that helps you align your model and the chatbot that you're building with your values. You explain to it what it shouldn't do, what it should do.

You make sure not to put sensitive data where it should- you shouldn't have. So you have a... You need to have multilayer approach.

In the end, it's a bit similar to a person. Let's say you hire an, a HR person, and they, they see all the salaries of people in the, in the company. How do you know that they don't tell it to someone from outside the company?

How do you know they don't do something bad? So you, you trust it, they know it's illegal, and you can hold them accountable because you can take them to court, to jail, like, so... But you can't really do these things with AI agents.

A-A-And, and I think that's the foundational issue, right? We want to give them abilities like humans have, so they can replace human tasks, but we can't really hold them accountable. And that's why we have all these security measures trying to, like, make it less and less and less probable that they'll do something that they shouldn't have.

But the moment you give them the ability, to do so, you should assume someone can convince the chatbot to do it. And I think that-that's the, th-that's the biggest security issue of today. Should I...

Do they build my agent HR, or should I not build it? Should I build a limited version of it? Um, exciting times.

It is an exciting time, and we're doing the best we can. I'll keep you fo-posted. Listen, this new, this new, offering from Mend, it's available now?

Yes, it's available now. Uh, and take a look at the website. Uh, it's super important, especially if you're building your own chatbots to your customers, and, like, security is something that we all need to do and have to do to keep the data of, like, of our users safe.

So if that's your situation, I encourage you looking at Mend, like, and see what we are doing. Uh, and even if not Mend, right? Look for open source, try to protect your AI.

Be cautious. Um, yes, that's my, of course, biased recommendation, but- I think that's good advice. No, no- Yeah.

Yeah ... I think that's excellent. Amit, be well, be safe, talk to us more, and, good luck with you and to all our friends at Mend.

Thank you, and thank you, Alan. Thank you. Was great.

All right. It was excellent. io here.

Go check out their new, system prompt hardening solution. It, it, you know, in today's world, it's something we need. We're gonna take a break here on Techstrong gang.

We'll be back in a minute. Hey everyone, it's Alan Schimmel from Techstrong. You know, we're gonna continue with this fantastic series we're doing of sessions between some of the leaders at Microsoft, as well as analysts from the Futurum Group.

In this next session, we're lucky to have Clay Wesner. Clay is the partner for GPM Power Apps Studios at Microsoft, and Futurum analyst, Keith Kirkpatrick. In today's, session, it's really a customer success story where Clay, joined by Keith, are gonna delve into a real-world customer story.

In this case, Wells Fargo offering a blueprint for leaders ready to scale success in the age of intelligent apps. You're gonna see how Plow- Power Platform is being used to modernize complex regulated workflows with Copilot Studio agents and Power Apps. This session will highlight architecture, business impact, and lessons learned from deploying intelligent apps at scale.

I think it's really a great session you're gonna enjoy. Let's go to Clay and Keith. And hi, I'm Keith Kirkpatrick, Research Director with the Futurum Group.

I cover enterprise software and digital workflows. Hi, my name's Clay Wesner. I look after our low-code developer experiences on the Power Platform.

Well, thanks for joining me today, Clay. Maybe, Clay, you could talk to me, though, a little bit how Power Platform can actually help these organizations balance that, that agility to handle these types of, of scenarios with their compliance needs that often come up when you're dealing with things like banking or insurance or, or any one of these regulated types of processes. Yeah, absolutely.

And it... You know, we, within the product, we sort of refer to this as managed platform because it is very much a feature of, of the platform of how you can govern at this scale. And, and this has come from, you know, not just, us deciding exactly what's gonna be in there, but really f-folks and customers leveraging low code over the last 10 years and evolving to have a really, really strong governance.

Because I, I think we learned very l-early on in the journey that if those guardrails are not there-People are just inclined to wanna turn it off. Um, and, and we're very much-- I use the, the word guardrails deliberately, and a lot of the things we do in a managed platform is focused around how do we give you the right control so you can still enable these types of tools, whether it be building apps, building automations out at scale, but do it in a way with the, the right sort of controls and guardrails on it. And so, like, examples are things like data loss prevention.

So I can set rules around what connectors and what data you can access versus someone else. And so I can also say how many people you can share an app or a workflow or something with. And so I can sort of mitigate the risk that you might be able to, to have working in low code versus someone that's received more training or onboarded to the platform.

And so typically what we see customers do is sort of implement this zoned approach of, you know, their, their zone one is everyone in the organization, and they say, "You can build apps for personal productivity, you can connect to your office data, you can sort of work with those well-known sources, and you can go and share apps and flows and agents with up to ten people," as an example. "But once you wanna go beyond that, we want you to engage a little more with IT. We wanna make sure things are supported.

" And then what typically happens is we'll then have a zone two, which is potentially some more sensitive data, potentially, you know, a, ability to share with more people within the organization. " And then that final zone would be your IT, your dev center, who's working with your, your really critical data around things like finance and, and HR. powerautomate and start building, and they're not going to fall into a trap there.

They're gonna fall into the pit of success because we've, we've put those right guardrails on what they can access and what they can do. If you look at not just if we're talking about, let's say, agentic technology, but just everything, if you look at the development of the smartphone, everyone expects sort of a consumer-grade experience throughout all facets of their life. And, and I guess that, you know, do you see that sort of pushing or helping to evolve kind of what cus-customer success might look like, you know, not just now, but into the future?

You know, again, earlier in the, the low-code journey, it was always IT departments, development teams that were looking at the low-code platform. And more and more these days, as we're talking to customers, it'll be their employee experience team. You know, it will be folks responsible for actually healthy and, and, and productive employee experiences.

And that's what I mean. It's not just about cost saving, but, you know, it's about bringing the right tools in. There's the SNCF, the French railway, are actually a really good example.

They run Power School, which is an onboarding school for the whole Power Platform when any new employee starts. And this is becoming a really, really common practice that more and more folks are, as they join an organization, they're getting training on these tools, not as something they have to use to do their job, but as a benefit to them to be able to do their job in a more productive way. And I think, again, the, the consumer push and acceleration of AI is, is just accelerating that within the enterprise as well.

Right. When you're talking about human in the loop, you raise a really good point, because ultimately, this is still new technology, and you wanna make sure that particularly in, you know, you're dealing in a commercial environment, that you don't want this agentic technology to sort of run wild or unchecked. So I'm just curious if you could talk a little bit about, have you seen other examples where customers have actually deployed their sort of checks, and, and balances to make sure that their technology does what it's supposed to?

Yeah, absolutely. And there, there's a couple of ways we're seeing folks doing that. One is just in how we define and build the agents and the tools themself.

While that agent has the ability to issue refunds, it can only do them up to a hundred pounds. So it has very specific guidelines built into it that once it goes over certain criteria, loop in a human, send them an approval workflow so that they can approve this, review the details. So that first one is just very structured, giving the agent details.

Um, the other side and, and this is where we've sort of really seen how apps have evolved in the last couple of years. If you've been looking at what we've done with Power Apps, we've introduced this concept of an agent feed, which is really about in the same UI that you would come into the app and, and do your day-to-day work, you start getting this feed of activity from the agents that are in your digital team effectively. And so you can start seeing where they're completing actions, where they might need assistance, or where they're getting blocked.

And so what we're starting to see there is even our UI patterns of what we traditionally thought an app was is starting to bring in this agentic behavior to give, you know what I mean, that human in a loop and that oversight capabilities. So I still want someone to have a really clear view of what tasks are being completed by the agents, what's being completed by AI. And in that view-Get-- be able to get into the reasoning, understand the logic and sort of the thought process that the agent followed behind it.

So it's not a mystery of why something progressed or why an action was performed, but as the, as the human responsible managing that team of agents, I can effectively go in and see why it did something that might be then come a teaching moment for the agent where we correct that behavior or change it for future cases as well. Well, you know, it-it's really interesting you mentioned sort of the generational shifts that are going on. You know, we're seeing the, you know, entry of these, I guess you'd call them AI natives, coming into the workforce where they don't know anything other than a world with AI.

Uh, and I guess, you know, that kind of begs the question, you know, we've heard so much about AI in the past, you know, particularly last couple of years. Can you talk to me a little bit about, you know, s- what role can AI actually play within customer success? Because, you know, it's a wide, you know, AI has so many capabilities, but I'd just be curious to see if we could boil it down to this function.

Yeah. I-- and I think it honestly depends on the customer and how they're approaching it. You know, one of my favorite examples of, I think, sort of scale and pace, PG&E here in the United States, they're a big power platform user, and we, we talk about scale.

I think they estimate since they started their journey in twenty twenty-one along the lines of like thirty-eight million dollars in savings that they accrue to the power platform, like huge- Yeah. -in, in, in terms of scale. Um, but so much of actually what they've implemented is not just, you know, cost efficiencies.

They introduced an agent called Peggy, and they actually have a nice little avatar for Peggy that they, they introduced across the organization. And Peggy now handles... it's between thirty to forty percent of their IT help desk calls.

So built in Copilot Studio, Peggy has access to their knowledge base, all their policies and documentation. And just Peggy one agent, they estimate saves them about eight hundred thousand dollars a year. And it's- Wow.

-it's absolutely transformational. And so even with the savings they were getting on the power platform between apps and automation, there is a limit. Mm-hmm.

There's a limit to how much productivity that that can drive. Agentic, you know what I mean, tools and, and, and what people are able to build in Copilot Studio has really just broken through that, that barrier. And, you know, you look at, again, someone like PG&E, when they implemented Peggy, it was very simple, looking over knowledge bases, access to information.

It helped a large volume of sort of tickets that would come through the help desk that used to be a human replying to an email or replying to an IM. Those humans now are actually providing much higher quality support on a, on more technical cases. They're not helping someone log into Citrix for the first time or, or point them to something that's really well documented.

Peggy's able to do that. But then they've also continued to evolve it over time. And so again, one of my, one of my favorites that Peggy can do is getting folks that get locked out of their SAP accounts.

One of the most common things that IT apparently happens thousands of times, and now Peggy using an integration between Copilot Studio and Power Automate can actually open up SAP and go and unblock that person's account for them after they interact with her on Teams. And so this was something that was critical to an end user to get unblocked really, really quickly. Peggy's able to do that for them fast.

But it wasn't high value from an IT support team and what they were really providing, them going and opening up an account and unchecking a blocked checkbox. And so I feel it's a really good example of where they started simple, they focused over sort of knowledge base examples, they evolved it into actions, but it's something where they've gone for a, a high volume, you know, cost inefficient area. They've applied agentic AI to it, and that's something that go back three or four years ago would've been an extremely expensive tool to go and implement.

Leveraging LLMs and leveraging Copilot Studio, they've been able to do all that in low code, which is super impressive. I, I'm curious, you know, one thing, Clay, that you alluded to earlier is if we think about how apps were pre-- you know, previously developed and rolled out, it was IT who kind of managed that. Now what it sounds like what you're saying is we're getting to the point where, you know, business leaders or even folks who are, are working within departments may be able to actually launch apps or launch agents obviously with that human in the loop and with those, you know, specific guardrails.

Are you seeing any kind of patterns emerging in terms of, you know, customers who've successfully scaled these, this agentic automation from more of a grassroots approach as opposed to springing from IT? Yeah, you're absolutely right. I mean, we sort of see an approach from both directions, and some, some customers very deliberately approach it from one or the other to start with.

I actually feel like all the examples I've, I've, I've sort of talked about today do quite well balancing both spectrums and both ends of the spectrum, sorry. And I think that's where you start getting the real value multipliers. PG&E, great example.

I talked about Peggy earlier. That's an IT or centrally led tool. It was about optimizing a process within the IT team.

But at the same time, they have thousands of developers across their organizations. Across their, across their team. And they've sort of very deliberately focused their center of excellence, their digital transformation team on a few core objectives.

So that's the team that sets their governance policies, makes sure it's scalable, and then they also support and train those different sort of divisional leads across the company. PG&E actually, again, I think they're on the spectrum, the end of the spectrum where they're doing this, you know, in a really amazing way. They have a conference every year called Level Up Now, where they actually get together all their citizen developers and, and those divisional leads from across the company to come together, share stories, share learnings, and sort of explain new technology.

But it starts becoming a real cultural tool in that they're enabling people to go and solve these problems, make themselves and their teams more efficient, and there's, there's reward that comes from that. You know, they're getting folks together, they're getting a lot of learning. And so I think, you know, while lots of companies are enabling citizen development, the ones where we see it's truly being successful, they're bringing this level of evangelism to it.

Well, Clay, maybe you can talk a little bit about some of these platform features that, that are kind of critical for managing customer success initiatives, because it really seems like, you know, it-- there-- obviously, you have the human component, but there's also the technology side in terms of making sure there are the right tools in place to help organizations, you know, address all of these issues. There's, there's obviously the human and technology component. I would also say there's just the practices and, and, and sort of learnings.

We actually have some good documented platform guidance out there of, like, what are the best practices in, in thinking about this zoned approach that I was talking about and in how people, can sort of apply different levels of control to different parts of the organization. I would say then we start looking at the specific technology. One, a lot of those guardrails just light up directly in the product.

So as a new citizen developer, as a maker, when I go land at any one of the power platform tools, I can get welcome guidance with links to internal learning, explanations of where I can go to support. I get routed to my own personal developer environment. So I actually have a sort of controlled, dedicated environment for me to go explore in, to experiment in.

I'm not sort of working in prod. Um, I, I have the ability to be controlled. And then things like pipelines, which effectively are a low-code ALM tool, so that once I do build something, I can either use it for my-myself and my personal environment, but if it gets to the point where it does make sense for it to be deployed somewhere centrally, leveraged by others, I can go through an automated deployment process where the right checks go.

I have an AI advisor that reviews my code, makes sure my apps are secure and performant and accessible, and then get the right approvals before that gets deployed. And it's really that mix of, you know, we wanna democratize, we wanna make these things available to everyone across the organization, but then have these right built-in tools so that you don't have to go read a wiki to find out what's the process that you should follow. It's built-in to the developer tool.

So I kind of just, as I start building, get guided to the right environment, I get guided to use the right data, I get guided to share it and deploy it in the right way. And all of that we bundle up and sort of leverage within that managed environment, which gives the, the admins, the IT, the central digital teams that control centrally to sort of set up those tools and that content that they want available across the organization. It sounds like all of these tools, you know, really underscore what you were talking about before, which is this culture of trying to utilize technology in a way where it's deployed at the right time, in the right space, and with the appropriate guardrails, but while still fostering a culture of experimentation and, and ensuring that people feel empowered to use these new tools.

You're absolutely right. Like, the cultural-- I think when we talk about your, your first question about, like, what's the new definition of customer success, you know, I think it's the customers that have implemented the right culture, and it feeling like it is a culture of empowerment and experimentation, not, you know what I mean, not something that they have to fight really hard to get access to. Because that's where a lot of these examples where we have customers turn around, they've built something that's ended up saving them millions of dollars, it came from the expert that was involved in the business process.

It didn't come from a central team. And to get that creativity and get that ideation, you need to give people access to these tools. " Uh, and I think, you know, so will users, so will makers.

They will find a way. And to restrict these tools, to, to hide them, folks will go find a tool on the web that can help them be more efficient in their job. The companies that are doing this right are making it part of their culture to provide those tools and just really enable people from the get-go.

The technology is probably going to be more accurate over time if you're talking about trying to, you know, really assess, you know, images and differences between them. But, you know, one of the other things I'm really curious about is how can agentic AI and, and all of this technology be used in regulated industries? I'm thinking in particular financial services, banking, insurance, where, you know, there's a lot of complex process, but you also have to be mindful of all of the, regulations that are, attached to those industries.

Yeah. It-And it's actually quite surprising, I think, in, in this technology shift with AI compared to when we moved to the cloud, compared to internet, compared to a lot of the others, I think actually the regulated industries have, have actually been quite a lot of the front runners on this. Um, you know, EY, for example, built PowerPost, which helped them with their financial processing, sort of end-of-month processing.

They built this as a, as a sort of typical low-code application. They're already looking at how they bring agentic checks into it to make sure that things are being posted into the right period, that they write... they have the right information.

Again, time-consuming sort of manual checks. Wells Fargo have rolled out agents to more than four thousand branches. You know what I mean?

A huge, huge number. And they targeted a process that was around their branch forms and procedure management. And this is something that was particularly time-consuming.

So if you went into a branch and said, "I need to set a power of attorney," or, "I need to open an account," and, you know, under a maybe so a non-traditional circumstance, there's a huge amount of internal documentation around those procedures, the right forms, the right information to collect. And before, that would mean as a customer is standing there with the branch member, they're looking up that information, trying to go find the right procedure, going in right-- going to find the right form. So a heavily regulated scenario, but also really impacting a customer who's literally standing in front of you waiting, you know, maybe on their lunch break, trying to, trying to get through the bank really quickly.

And so they rolled out an agent, you know, across all their branches to actually manage that forms and procedure scenarios. And so that now in the branches, those employees are jumping straight onto an agent, talking about the scenario that the customer has, and working with this agentic AI to basically get guidance on the right forms, the right procedures to follow. Even in these regulated industries, they're seeing the value in AI, and I think it's more about how they do it, making sure they have the right checks in place, making sure they have the right guardrails rather than what they probably would have done five years ago, where they just tried to turn it off.

You know, we, we talked a little bit about, you know, potential friction there, but are there any other sort of potential hurdles that organizations need to be wary of, and, and what, what's sort of your take on a solution? Like most things, we talked about human in the loop. You know, making sure you introduce this technology in the right way to organizations is really, really important.

I mentioned EY earlier. They were really, really successful in after building PowerPost, which helps them manage their, their sort of end-of-month financial processing. It simplified it.

It brought in some agentic behavior to vali-validate quality, and it-- they had, like, huge gains in efficiencies in both. I think it was seventy percent in, in sort of the time, or ninety-five percent in lead time to get things posted and about a thirty-five percent cost saving for them. So like real, real sort of impact to the efficiencies of their users.

But what they did really well was once they built that tool, they told that story, they evangelized it. And so they helped people understand that this is how this technology was helping them. This is how it was implemented.

And that not only made, obviously, people a lot more receptive to onboard and leverage the technology, but it also started driving this ideation of other things to go improve within the organization and using similar technology. A lot of these companies are not coming in and doing a full low-code approach of apps and agents and automation and reports all on day one. Where we're seeing folks be really successful is they're leveraging the composability of the platform.

You know, they're starting with, for example, they might have a, a legacy application that's inefficient for a user. So they go and use an app, they build more efficient, streamlined UI over the top of that. That's an incremental solution they can deploy, they can get out to their users and start seeing benefits.

Then on that same app, they can go and add automation. They can start getting approval workflows. Then they can start bringing in agentic AI, getting that automation and that AI behavior incrementally building these solutions over time.

And it's very much, you know, intentionally how we've designed the platform in that these are not all or nothing solutions. And, you know, back to our earlier conversation, pace is extremely important these days, and people don't wanna go do a twelve-month waterfall project of every requirement met. They wanna find ways to incrementally build.

And by leveraging a platform that has common governance, these tools are designed to work together, apps with automation, with agentic behavior integrated into Copilot with that unified platform. So essentially, you're setting up a framework to enable organizations to really drive these best practices in terms of making sure that, yes, you are implementing new technology, but you're doing it in a thoughtful way where you have the right checks in place and, you know, you really are making sure there's, you know, other things that, that you need there. You need the audit trails.

You need to make sure that, you know, when you do a project, you're going back and you're actually assessing, you know, does the technology achieve the goals that we set out to, set out to when we deployed it. E-exactly. And I think it's that, you know, there's two parts to it.

One is that being proactive. So as you're releasing a new app or a new agent to the organization, do you have the right controls around it, the right guardrails from the beginning? And again, our goal is let's have the right framework, the right tools, the right guidance to go really enable that and let an organizationTailor those guardrails to sort of accommodate their level of risk, what they're, they're comfortable with doing.

But then on the flip is make sure we just have the right visibility, the right auditability, so that as you're leveraging AI more and more within the organization, it's really transparent- Mm-hmm ... about what it's doing. You know, I think one of my favorite things with, Copilot Studio, and Pets at Home is a great example of this as it's interacting with customers on customer service.

You can go into any step through any sort of run or action the agent has performed and see-- understand its thought process. Why did it do this particular step? What were the inputs?

What were the outputs? What were the reasoning? Um, and not just understand it, but then also help teach it for-- to handle sort of moments, at a different way in the future.

And I think having those, those sort of tools from a governance perspective just built in, again, you know, we talk about it being unified for the developer, unified for the end user, but also for the, for the admin, so that they're doing in this sort of a central and controlled way. And even then, whether you're building an app, an automation, an agent, you know, you've got that composability across the platform. But I don't think admins really want a super composable admin story.

They, they want that to be a lot more unified and, and controlled. So, you know, it-it's bringing the, the blend of those worlds of let's bring together multiple technology, multiple tools, but make sure then you sort of have one central view of, of how it's all coming together. If you want to really drive the use of new technology, you need to do it in a very stepwise fashion using a platform that allows you to unify people, processes, technology.

It doesn't make any sense to try to do it in a very disjointed way. You won't have the governance required to do it safely. You'll confuse people in terms of which tool should I use?

Which approach should I use? Ultimately, it really does matter to make sure that you have a unified way of approaching the implement-implementation of new technology. It's also really critical to make sure that as you go about your journey, whether it's implementing low-code processes, implementing agentic technology, to have a clear understanding of your business goals.

What outcomes do you want? How are you going to measure them? And then how are you going to take all of these different learnings and then streamline it so you can actually apply it and scale it over the enterprise, not just for today, not just for tomorrow, but well into the future.

And finally, I think the most important thing that kind of resonated with me today is you need to look for a trusted partner, trusted technology partner to help you through this journey. Agentic technology is very new. Low code, yes, it's been around for a while, but, you know, there are still quite a few pitfalls that can be out there.

To, you know, to go it on your own can be very, very challenging because you have all of that risk of potentially opening yourself up for errors, missteps, and of course, there's that, you know, we talked about it a little bit today, regulatory concerns. It makes a lot of sense to tr- to partner with a company that has experience w-with other enterprises to deliver these types of benefits using that new technology. Hello and welcome.

So glad you've joined us for the third part in our series around adopting AI, using AI in the mainframe environment. My name is Mitch Ashley, and I lead the software lifecycle engineering practice with the Futurum Group. Now we're gonna be talking about acting with confidence.

We're using AI as a change agent or agent of change, if you will. And I have a great guest with me talking today about this topic, Anthony DiSarro. Anthony is Senior Director of Architecture for AI with BMC Software, and our series is, presented or sponsored by BMC Software.

Thank them very much for putting this together. Anthony, welcome. Mitch, thanks for having me.

It's been great. The f-the first two segments we've gotten to do together has been really fantastic. So I'd really love to hear more about kind of AI agents or agentic AI and how this differs from the traditional AI, generative AI type solutions we've been talking about.

Absolutely, and this is my favorite subject of late. I've been spending a lot of time in the space around AI agents and agentic AI. Basically, it comes down to autonomy with, with AI.

So in our first two segments, we talked about AI as an advisor and AI as a partner in the second segment. Now, great conversation around that, but there's one fundamental challenge with both of those perspectives on AI. It's a reactive model, Mitch, right?

The AI is waiting for you and I to go interact with it, whether that's a chat experience or even it's infused in a product experience. It's sitting there waiting for us to go interact with it. This is wonderful technology.

This is great technology. We don't want it sitting idle twenty-four by seven waiting for us to approach it. We want this technology to work for us twenty-four by seven.

So this is what I like to talk to our customers about is this is a shift from reactive AI to proactive AI with AI agents, where we can have agents that are running, again, twenty-four by seven Working on a task, working on achieving a goal, these agents would observe the data in their environment, they reason over this data to make decisions on what they should be doing, and then this is what sets it apart. Now, the agents or the AI acts upon that decision that it makes. It takes an action.

And the best part of that is after that sequence, the agent can learn from that experience and get better at its craft or at its work that it's, it's trying to do. So agents would... could exist in, you know, different organizations, and over time those agents would learn and grow to that particular environment.

This is a game changer when it comes to the AI space. Now, what does that, what does that look like in a mainframe environment? Could you give some examples of things you might have an agent tasks that might perform, things that it might do for you?

Oh, yeah. Absolutely. So we-- in all the previous segments, we talked a lot about let's go back to the COBOL, example that we had.

The COBOL code does, you know, code explain, for example, or explain a given situation i-in your ops space. Well, let's go back and look at the COBOL example. Instead of just explaining the code, maybe now the, you know, the AI agents can get to a point where they could do a code review.

They could make recommendations on how to improve the code. The, the AI agent could actually fix the code itself, right? So this is where the agents of change come in, where ag-uh, agents of action come in.

So it's no longer just giving you information. It can go out there, look at the code, improve the code for you, do a pull request, and get it back in-in-into the system all on its own without potentially human involvement in that. But we are nowhere near that part of total autonomy with AI agents.

Our approach and the approach that we see a lot going on in industry today, Mitch, is keeping that human in the loop. Why we wanna keep that human in the loop? We wanna keep controls and guardrails around the AI and its decision-making so that we understand what the AI is gonna do, and then maybe you hit that button in the user experience, and you tell the AI, "Yep, I agree with what you're going to do.

" And you could say, "Yes, go off and do that. " So the way you wanna look at this and the way we're looking at this is AI agents as actual digital workers, and orchestrate these digital workers just like you would your staff within your organization. You're gonna give these AI agents very focused responsibilities and jobs to do, and they will get better at that job over time with that learning capability I mentioned before.

So it's that taking that next step. And I, I'll give you just a real quick example. So I created an application about three weeks ago, and I had three AI agents that I created that were with me in my development environment, and we created this whole app over a four-day period, and I did not write one line of code.

I didn't write any code to, generate test cases, run those, unit tests. I didn't write any code to deploy it to containers and into my containerized environment. The AI agents did that all for me, and I was just orchestrating them just like I would an engineering team over my career, and it was fabulous experience.

It really is pretty amazing to see- Yes ... see AI work and doing so much of that for us. Yeah.

I think in my own experiences, Anthony, of code bases where code had... part of it hasn't been touched for a long time. Folks that may have been-- done that have long gone, and everyone's got reticent about even going into it and trying to make any changes, more or less understand it.

Think about the power, not just of freeing us up to do other tasks, but also going back and doing more with the code that we have, the things that we may have, you know, not wanted to touch for a long time. Yeah. Uh, I'd love to hear your perspective about, this isn't an all or nothing when we talk about agents or agentic AI.

It isn't turn everything over. You mentioned about, "Let me review your work. Okay.

" Talk about that. We've talked about human in the loop. Talk about that process of how we judge.

We start to use AI more and more to do agentic processes, but we still have control over what's happening. Yeah, yeah, yeah. A-a-absolutely.

You know, i-it's based on the technology itself today and the state that we're in. Even though we've made tremendous strides over the past two years, we're still at a very early stage. And, you know, AI in every agentic system that you use today, you, you see the messages that come out that AI could make mistakes.

You read-- You should double-check what AI is telling you, et cetera. And really that's what, what it's all about. So even the example where I, I built this app over this four-day period with AI agents, the AI agents didn't do anything without my approval.

So when it... If it, it generated a script or i-it generated a module or whatever I was directing it to do, I'm still a very seasoned engineer. I was a...

You know, I sit there. I study the code that it generated or the actio- the things that it wants to do in my environment, and I am giving that approval to, to, to move ahead with that action. That's where we are today.

That's where we need to be in the, in the near future. So when we build AI agents and we look at use cases for our agentic system, that's what we're looking at to address, is how do we have these agents that are working twenty-four by seven on some use case-But how do we connect that human back into the loop for that oversight that needs to, to happen? You almost wanna look at the AI agents as a junior staff member, right?

" You, you and I today would, would, would, would, would review that, right? That, that, that young develop- developer or that young ops person that's new in our space, we would s- you know, r- review what they're proposing to do, and then we would give them the green light to go do it or not, or, or we would have an exchange with them on how to make it better and improvement... improve it based on our experiences.

It's the same thing with AI agents today. You give the green light when you want the agent to go off and do that based on your satisfaction with what it wants to do, and then you could also then guide that agent on how to make better decisions over time. It's a real fascinating area that we're getting i- into.

I'm telling you, Mitch, the, the use of agents in an agentic AI system is gonna completely change how we build software systems moving into the future. We are just scratching the surface on this. Usually, you know, from, from a, a BMC, any portfolio perspective, w- we deliver a lot of functions and features through our product experiences across the breadth of our, our portfolio.

Imagine a future where BMC Software delivers solutions as a collection of agents that get deployed and then augment our existing solutions or provide additional insights to you, but it gets delivered as agents. These agents then could be reused amongst many different use cases, right? So you know what, back in the day when we got into this whole mindset of reusable services, right?

And service-oriented architecture, it's like you wanna create the services to do specific jobs, and then we tie all these services together to build a system. It's the same mindset with these... with, with AI agents.

It's like you... we're gonna create this pool of AI agents, but then we're gonna start to mix and match these agents to come up with use cases that were unimaginable just a few short years ago. You know, it, it has drastically changed our business already.

We are... We create software totally different than we would have six, 12 months ago. It's really pretty amazing.

You know, you... as you said, the word guide, guidance, I think is really kind of the human in the loop. It isn't just approval.

It is, yes, I'm guiding where, where, where AI is going for me, what the agent is doing. Now, there's some other things that I think you need to be have in place, right? When we talk about trust in what the a- agent is doing.

So you wanna know what it did and how it did it and what happened. Yes. What are those things that you need to have in place to be able to get that information, to have that oversight and insight- Yep ...

to what's happening? Yeah, yeah. So, and, we talked about, having that transparency or that observability level in your ag- in your AI system.

Well, that pushes right down into the AI agents, right? An AI agent, first of all, when we create an A- AI agent for a specific goal or task to achieve, there are things that we do when we build the AI system to make sure we ground that agent to only do that type of work that it was set out to do, and there's different AI techniques that we, we use to do that. But at, at the end of the day, that, that agent itself is totally audit- auditable, traceable, all the observability, it's all there in, in the, in that individual agent that we're, we're creating.

So all of our agents that we create part of our agentic system inherit this. The other thing that we could do is, and mentioned that we can infuse into BMC AMY Assistant, additional knowledge from the LLM. Well, now with AI agents, to keep them grounded and guardrailed to do the jobs they want, we can infuse into the AI system specific pieces of knowledge that is targeted for a particular agent in that system.

So the agent could use-- will, will use a language model. The agent will use knowledge base information, whether that's provided by the customer or provided by us, targeted to that particular agent. Everything to keep that agent grounded and/or fenced to do just that task or, or, or goal that it was designed to do, and again, totally auditable.

Um, so you can get into, like, the admin console for BMC AMY Assistant. You'll be able to explore all the agents that we have, and you'll be able to get all that observability information for those agents in the same, in, in the same way you would for an entire agentic system. Yeah.

I like to think of it as we don't have to just have free-range agents where they're operating on information and data that we provided. There's the concept of artifacts, right? This is...

These are sort of the, the rules of the road. Here's how we operate or what things we do when we use certain terms in our prompts. There's also...

That becomes the ca- canonical knowledge that you're using to set those guardrails and processes. Talk about how you take... You know, we're learning how to use AI too while we're implementing agents and new technology, and that can become part of this long-term expertise and knowledge that we're creating through using AI.

100% I, agree with, with that segment. Yes, absolutely. It's so transformative in what's gonna happen with AI agents.

It's... The level of trust will come with AI agents over time, and-We're no-- I just want s-say this, that we're nowhere near AI agents replacing a human being. The-- It's really kinda disturbing to read that and when, when, when folks are talking about that.

We're, we're, we're so far a-away from that when it come, when it comes to AI. The way we should be looking at AI agents is digital workers that make you and I better at our job day in and day out, and make our solutions better day in and day out, deliver higher value to our customers day in and day out. That's the way we should be looking at AI agents as we move forward on this journey.

Very good. Thoughts on next steps for folks that wanna move into this more agent, agentic type process. I'm sure it's not jump into the deep end of the pool, right?

There's a, there's a stepwise, approach that you can take to taking on certain kinds of work, what kinds of risks you wanna try, take on, et cetera. Yeah. So when you, when you start looking at agents that you wanna build out for your a-agentic system, you know, start with agents in that advisory role, right?

One of the things... A-and why I'm heading this way, one of the things that we learned is when you treat your AI system as a monolith, where it's trying to handle all kinds of requests coming across, say, your, your, your, your, your portfolio. Just like each one of, every one of us, if we're overtaxed doing many, many different things, we're not really good at any one particular thing.

But with AI agents, where you can give them specific language models to use, specific knowledge to use, specific guidance to use on what their job and role is, all of a sudden the responses you're getting from the AI system are really accurate and really good. So even if you're getting into the realm of l-let's just start with the advisor part of it, start breaking down your areas of advisement into agents, different agents, different advisor agents, and you'll see how much better those responses are from the AI system when you can focus those a, th-th-those, agents around its advisement responsibility. Start there.

That, that's a really, really, really good starting point. When you wanna start moving into the actionability part of that, and it doesn't have to be actionability a-as far as, you know, dealing with something tangible that puts your production environment at risk. When we talk about an AI agent taking an action, it could be simply observing a situation and sending Mitch, sending you an email and just bringing something to your attention.

That's actionability. But it's just sending an email or sending a message to a t- a Teams channel or whatever it may be. You could have those, those levels of actionability.

It doesn't have to be things like, you know, IPL, I think for... or something, right? Start killing processes.

No, please don't. Right. Starting jobs and things of that nature.

No, we're not, we're not talking about that. So advisement, observing a situation, and maybe just sending emails and notifications. So start with the notifications.

That will start building, again, the trust with agents and start building you do- it will start moving you down the road towards an agentic AI system. And as the technology that we're using evolves over time, our agents will evolve with that technology. Our agents will learn and grow within the environment they're running.

Our agents will learn and grow with each other. 'Cause we've been-- right now, up to this point, we've been talking about individual agents doing a particular task. But in a true and agentic system, this is where you could actually orchestrate more than one agent to work together as a workforce or a digital team to solve complex multi-step IT problems as an example.

You would have multiple agents working together to solve problems. And then when you follow that with using open industry standards, like the MCP, the Model Context Protocol, to connect agents to real-time systems or agent-to-agent protocol to get agents to communicate with each other, we get to a point where my organization, I'm creating AI agents, you're creating AI agents, all of a sudden we get to a point where we can collaborate on use cases together. You provide certain amount of agents, I provide a certain amount of agents, they work together and collaborate to solve problems.

So I see us at BMC working very closely with our customers in this journey because our customers too will be rolling out agentic AI systems in their environment. And I can see us working with our customers and collaborating on use cases with our collective agents together. So it's gonna open up lots of opportunity, open up lots of doors for options, as we go forward.

Very exciting. This is, this is the game changer. This is the transformative part of AI in the enterprise.

What a great time to be in technology. It really is. It is exciting and fun.

You know- And all on the mainframe, right? Yeah, all on the mainframe. You know, it, it is-- I, I love the way that this series came together, Anthony, and appreciate, your, your insights and also folks like Eric O'Dell and the team at-BMC Software that helped put this together, thinking about this process, this journey you go through of using it as an advisor, learning what you need to learn to move on to the next step, to use it as a partner, AI as a partner, and then into more the operator, the agent.

And I think, th-this is a really good series, I think, that will help people along as they get to that place in their journey, like, "Okay, I get what they were talking about. " So, thanks again. We appreciate all of your knowledge and your insights, and we hope everybody has gotten a lot out of this series.

Thank you for joining us, everyone. Thank you to BMC Software for having the foresight to put this together and help all of us on this journey that we're on together, and that's, that's the important part. We're in this together.

So we'll see you next time, hopefully on another series, on another conversation about AI adoption and use in the mainframe environment. Take care, everybody. Progressive CIOs are looking to help their organizations roll out AI in a responsible way to enhance productivity.

With so many AI-powered tools available and changing every day, it can be challenging for technology leaders to decide which to deploy. But the real issue is socializing this change with the people that will use the tools. In this episode of Utilizing AI from the Futurum Group, Siriu Mesheghian from, Barracuda Networks joins Olivier Blanchard, Jon Swartz, and myself to discuss how this well-known cybersecurity vendor has adopted AI.

Welcome to Utilizing AI, the podcast focused on practical applications of artificial intelligence from the Futurum Group. Every Wednesday, we explore news and use cases of the ways in which AI is transforming enterprise IT and the industries it serves. I'm your host, Stephen Foskett, President of the TechFieldDay business unit here at the Futurum Group.

Before we dive into this discussion, let's meet who's on the panel today. Yes, good morning. Thank you for having me.

I'm Siriu Mesheghian. I'm the CIO at Barracuda. And what I'm seeing across the enterprise is that the conversation's moving from which AI tools should we try to how do we run AI every day?

So that's securely, responsibly, and at scale. And for us at Barracuda, the unlock is combining experimentation with the right kind of guardrails, governance, and that's so it doesn't slow down innovation, but it makes it safe for everybody to operationalize the use of AI. Um, and because Barracuda obviously lives in cybersecurity, we're specifically focused on how AI changes both the threat landscape and the way that we defend, while also reshaping how people work day to day.

Hey, I'm Jon Swartz. I'm with, Textron Group. Although technically now, as of March first, I'm with the Futurum Group.

We merged with them. I'm the, senior content writer. I write a lot about AI.

I also write for a section on IT and DevOps and oc-occasionally security. I'm based in the Bay Area, and I'm happy to be here. And I'm Olivier Blanchard, research director with the Futurum Group and Futurum Research.

Uh, and my focus is intelligent devices. So it's, it's kind of like the edge side of the AI equation, where AI plugs into all the devices that are around us, and how AI spreads to the edge and, and, and populates the world around us as opposed to just being in the data center. So I'm happy to be here and talk about this.

Great, and, thank you so much for joining us, Siriu. Now, you're somebody who I have seen, you, you speak, you do podcasts, and so on, and you are very much on the, operational side of things in terms of helping your company, but also other people in the industry learn how to use AI to be productive, but also in a responsible way. And I think that that's really where we should start this conversation.

Um, you know, so how are you thinking about this whole idea of transitioning from AI as more of an experiment to something that is part of everyday operations? Yeah. So it's, it's definitely interesting taking it from being what has been an experimental state to, running it day to day.

Um, so we, you know, initially rolled out AI sort of, you know, gently and slowly. We rolled out some chatbot, infrastructure. We helped people get comfortable with the concepts.

We wanted to meet people where we-- where they were. Um, Barracuda is a company that produces software, so we have people that are in product and engineering. Those folks are very comfortable with concepts of AI.

But we also have employees who are not comfortable at all with those concepts, and it's important for us to recognize that. So we started off, as I said, meeting people where they were. So we, decided that we wanted to instill a, a program where employees were encouraged to get educational training.

So we offered that to everybody, no matter who they were. We had, personas of types of training for them. Uh, so operationalizing it meant first kind of getting people comfortable with those concepts and then deciding how we were gonna roll it out.

So we've got obviously chatbots for people who are more comfortable with that, but then we've got ideas where we're building out agents and then looking at how we're building that portfolio, so it becomes, a way that we are structured and moving through, all of the projects that we have. And in terms of that, we, we're not just building things because people are interested in it. We wanna make sure that we are very, um-Thoughtful in the approach.

And if we have to hit the quit button because something's not getting to production fast enough, we're doing that. Uh, but we're also making sure that we have guardrails for people to test all of the AI technology that we think is important for them. And then y-you know, an undertone to that is also cost.

So we wanna make sure that we're being careful about how much we're spending on it. Um, while that isn't something that we want people to be very focused on, I myself am. So that's how we're thinking about it.

Hey, can I ask a, a qu- just kinda a basic question? I get a lot of, email studies, and there's a common refrain or a common, narrative that, that's been emerging or has been in effect for a couple of, a couple of months, and I think Olivia and I have come across this a lot, and it's this, this idea of AI pilots, rarely scaling or trying to gain traction. I'm wondering, Ciria, what are the most significant challenges organizations face now?

Is it the technology itself or, or something else that, that crops up? I think it used to be the technology. It's no longer that.

The, the challenge I'm seeing is in the operating model around it. So data governance, the risk controls, the decision rights, preventing shadow I- AI. I was gonna say shadow IT, because that rolls off my tongue so much easier.

We've been saying that for so many years. But shadow AI is the name of the game more, more recently. Um, shadow AI from showing up faster than organizations like ours can see it and manage it.

So because AI is running on your most valuable asset, your data, the question becomes: how do you protect that sensitive data and the flows without freezing innovation? Um, that balance is the harder part. Not spinning up the model, not finding the tool, because the tools are falling out, out of the sky left and right.

That's no longer the issue. Um, it's, it's not buying the tool, it's trying to figure out the operation of it and using it within safe guardrails. Yeah, so on that, this is, this is not gonna be a, a conversation about technology.

It's, it's really a conversation about change management, right? Um- That's right, yeah. And so having, having been part of, of many efforts, that, that touched on change management from digital transformation, you know, one to digital transformation two, digital transformation three, which I feel like this is part of, how do you develop...

First of all, how do you develop best practices in, in your organization? And with a technology that moves so quickly, and again, with this sort of like waterfall of potential products and solutions that you could, you could insert into your organization, all of the, testing that has to be done, all of the, the, the fireproofing, all the training, how do you build a, a comprehensive living, breathing, practice of change management that is focused on AI integration? That is a big question.

Um- There are a lot of themes to that. Building a change management focus around AI, it's, it's probably the most important function that you could have. So a change management practice that we have going on, kind of has started with an AI council that we, we built from a very small state about a year and a half ago or more.

It started off with just a small group of people, myself, it, it had legal and compliance, so just a handful of people. We were really focused on some of the, the, the compliance aspects of AI at first. We weren't really focused on much more than that.

Uh, that and having an AI policy that we wanted to, instill in our environment and make sure that that was a living, breathing document that we could socialize easily and well. What's happened since then is that group of people has grown substantially, down to the point that we can't manage it well, but it's grown in a structure where we have a couple of important subcommittees. We've got a go-to-market subcommittee.

We also have a product and engineering subcommittee. Two really important areas that deserve a time of their own. And then we've got a core group that has representation from every executive pillar in the company.

And so change management is rooted there. That's where we're talking about not only the compliance policies that we need to be mindful of, but we're also talking about the policies of how we roll out AI in general. How is it that we're going to make it more f- frictionless for people to look at tools?

How is it that we're gonna protect our data and keep that more for ourselves to manage rather than kind of, you know, exacting that force on people who we're trying to create experimental vibes and fun for? Uh, so that's, that's more about how we're thinking about change management. But we're sharing it amongst everyone.

The, the idea of changing AI and having that be a paradigm shift within our culture is something that is not just simply an IT experiment. It's not up to me alone. Sure, I can be, an ambassador of it, but I am not the one responsible.

It's me who is sharing the concepts with other executives that work at Barracuda and helping us all make sure that from the top down and the bottom up, the message is shared and that we're all ensuring that our culture is one of people-focused first, empowered by AI, and that people feel comfortable and confident about using the tools and the methodologies that are available to them. You know, I think Steven is gonna ask you about people and about upskilling. But before, before we get to that, I wanted to ask you what you've done, and, you mentioned it earlier, at the organizational level to ensure AI isn't just being used, but being used efficiently and safely, and ultimately there's a certain responsibility, I guess, ultimately, I think you're part of that, of who's responsible for the success or failure of the initiatives at the company.

But can you mention some of the things internally at Barracuda, and then maybe if you even wanna expand on that, what other companies are doing or what you're seeing other companies are doing to, to ensure AI is being used efficiently and safely? Yeah. So the AI council is, is the most important structure that we have to ensure that AI is being used safely and effectively.

Making sure that our AI policy is socialized is something that is another aspect of it. So it's not something that we rolled out a couple of years ago, and we expect people to understand. It's something that we remind our employees of as regularly as we think it's important without being a big drag, so there's that.

Um, we also have policies to make sure that we, explain how it is that we will use tools efficiently and the way that we'll be able to experiment with them. So, you know, we have intake systems for the way that we, we want to initiate new, new tools and SaaS and all of that. But AI kind of gets, ahead of the pack in the way that we investigate and look at, you know, proofs of concept for those types of tools because we know that they're important, and we know that they're a differentiating factor for a lot of the ways that we may work on things.

Um, so processes around all of that are important, making sure that they're documented and available for people to see. Can I ask you a quick, just to interject something. What's the res-- how receptive are your employees to using AI?

I know you're a tech company, so I'm, I'm, I'm assuming they're much more open to the idea, but there seems to be a broad kind of fear, dread, paranoia among the most of the workforce. There-- We have all walks of life at Barracuda, and, and I just came back from a Gartner conference, a CIO conference, and I validated the same, with many of my peers. Doesn't matter if you are a tech company, you have employees that are working in core business units that may have been at your company, depending on its age, for years, and they're doing things very manually, and AI seems like a foreign concept to them.

So I wanna focus on that group of employees for a moment. They are very AI-averse, and they're very concerned about their jobs. They believe that using AI is something that will, potentially replace their role, and it's up to us to make sure that we show them that if we are using AI for their benefit, that we show them that, you know, we're not working to replace their role.

We're looking to up-level their skills, that we wanna make them more efficient, that we want them to stop doing things manually, to think about a new way of working. Uh, so those are really important concepts to ensure that we've got filtered through the environment, so people understand that we're not out there just, y-you know, trying to scrape, through and to, and prune out certain roles. That's not what we're up to.

Uh, and then we've got more, you know, precocious, smarter with AI employees who, are out there looking for the latest and greatest tool, and they're not afraid. They're, they're excited about it. They're the ones who are the ambassadors of, of use.

But we also have, and what I'm trying to lead into is this concept of AI champion, where we have people sprinkled through the environment that, in, in these groups where you have employees that are more concerned about AI, people who have a, a stronger vision that can help them grow comfort. And then we've got, all kinds of activities like AI hackathons that are available for every single type of employee. It doesn't matter if the employee works in product and engineering or if they work in finance or analytics or HR, they're all invited to participate.

And, just so you know, the winners of some of those AI hackathons happen to be not from product and engineering, but from some of the other business groups, and it's very exciting to see what some of the, their ideas are. Yeah, this is a really interesting approach. Um, and well, first off, I mean, you really sound like, like a CIO here in terms of having this be about technology, but also about people.

And this is really a-an exciting and interesting approach because, you know, on the one hand, you are, talking about a progressive vision of AI, of adopting AI tools. But on the other hand, you're aware of the fact that not everyone is necessarily on board with that, and you have to, you know, your job is also to help them move forward and help them adopt the technology to be more productive. Uh, when you're doing this, when you're out there with the people, is, is the major challenge for the company about trying to help upskill existing workers, or is it a bigger challenge to find people who have these skills and bring them into the organization?

You know, it's, it's interesting to think about the assets that you have in your company today. They're the ones that have the tribal knowledge of what's been happening. Again, depending on how old your company is, what the vintage is of it, you may have employees that have been there a very long time.

They know where all the things are, the way things operate. They're the, the oftentimes the, the ones that are working in the core business units that are doing things manually, just talked about those folks, the ones that are more concerned about AI. Those are the ones that I want to learn aboutThe concepts of AI and how it can work for them the most.

I, I definitely want to create a strong core team of AI-adopting engineers who can help get people over the hurdle and get them comfortable. But never looking to bring too many people in from the outside to replace all the people who have the heart and soul of Barracuda, because that's what's made us so great. Like I said earlier, you know, we're people-powered and we're enhanced and improved by AI, and I fully believe that to be true.

So what we've been very focused on is making sure the people we have in our company have easy access to the tools that are familiar, that are, you know, the ones off the shelf they see talked about on commercials every day, available to them to try. And we also have tools that are much more advanced for those people who feel very comfortable and confident with those. And then we have these AI champions to help the ones that are feeling more, you know, like AI is more risky for them.

So we have AI champions, we have the core team of engineers who are there to help get them over this concern, and that's how we're trying to approach the workforce, who some of whom are, you know, like lower adopters, and then some of whom are much stronger. So about that, that-- those are really good answers. I'm, I'm kind of, always focused on ROI, uh- Yes ...

just because obviously with AI we're talking about efficiency, operational and otherwise, but there's also a cost associated with that. And then the fact that you have to sort of relearn or like learn how to do things differently, even if at the end you're gonna be more efficient with your work and you're going to be able to do more things with agents. Um, initially there is, there's a little bit of friction, there's a little bit of cost there.

So I'm wondering, the broad question is how do you measure, success? But, the more specific question is, how do you measure success now versus how do you think you'll be measuring success six months to twelve months from now with regards to your- Hmm ... your AI practice?

I love that question. So today we, we run AI like a product, like a program. It's really important to me that we do that, that it's not just sort of splattered around all over the place that, you know, this group has some AI requests that they have that we're kind of helping around with, and then another group the same.

I have a full organized portfolio of AI initiatives that's organized by executive pillar, and, and I have in there metrics of success, so I can see exactly what it is that we're working on and how it is that it's gonna benefit either that group, our company, will it be time savings? Will it, you know, like what exactly will the benefit be? " We can't do all of it.

Our portfolio is very large. We've had many success stories. We're doing many cool things, but we also have a lot of backlog, and I wanna make sure that we're choosing the right items to focus on that are going to have the biggest impact.

So what's going to make the biggest difference now? And then, Olivier, to your question about how we're measuring today versus how we'll measure tomorrow, I'd only like to say that we'll take the framework that we're building today and further improve that. So we'll iterate on it, so we are gathering better dimensions of information so that we can make better decisions and cut bait earlier.

Maybe we can a, you know, like figure out earlier on in the process of something that we're working on that may not be the right idea, and we'll be able to change tack much quicker. And, and that's money savings right there, because you're right, it's an expensive endeavor. So we, we do want to ensure that we're focused on the right things.

And, and when you say expense, it's not just the tool itself, it's not just the cycles, it's, it's people that are involved. It's the humans that have to avail themselves to tell you what their processes are. If you're not process mining on their machine, it could be because the processes happen outside of their machine.

You can't actually process mine on it. You can't do desktop procedural mapping because everything happens, like they're walking down the hallway. You know, like sometimes you can't do that.

Uh, you know, so it's, it's a very interesting journey to see how it all kind of plays out. But to your question, I would like to say that we get even better at figuring out how to map all of this and to capture the success. Yeah.

So I'm, I'm gonna ask you a-about, governance frameworks and, and how do they stay flexible enough to handle different international regulations or executive mandates? I know- ... the State Department came out with a memo that was leaked around resistance to GDPR and other things in Europe, but I'm wondering if you could maybe stre- answer that, this, this, this handling of governance frameworks and, and navigating them.

Yeah. So that's excellent question because AI is something that needs governance, and it's not just something that we can figure out for ourselves. We should all have the DNA to want to protect- Well, you know, our data and, and also want to figure out how it is that we ensure that AI is, is not going to do things that wind up being harmful.

But, but we do need to rely on the, you know, structures that are larger than ours, and oftentimes that winds up being govern- you know, the, the governments that are going to be, initiating some of these frameworks that are going to help us keep AI safe. But how do we... You know, to your question, the way that I see it is that I'm a big fan of kind of principle-based governance, with these region-specific controls.

So m-meaning that every area of the world is going to have some sort of different flavor of AI governance. Um, so it's, you know, principle-based governance, kind of, you know, like the way that you would think about the-- it principally across everything and then, you know, regionally, how you'd wanna think about your controls. So the core doesn't change.

Uh, you-you've got your, your risk-based controls and your, your core principles that are protecting the way that you operate, especially if you're a cybersecurity company. And that, that is, you know, no holds barred. That's exactly how you're gonna-- you're going to operate.

And then you adapt your implementation details, by jurisdiction and by mandate. Um, and that's why we anchor on these internationally recognized approaches. So, you know, like NIST AI and, and things like that.

So that we, you know, we can apply this differently, and, and because governments are rolling out AI mandates very frequently, we do need to be very flexible in the way that we can approach all of these and, and make sure that we aren't doing, you know, that we can m-maintain this flexibility as we see things, change across the world. Yeah. So, follow-up to that.

What are your-- Currently, what are your two or three biggest friction points? Is it the adoption? Is it the resistance?

Is it the selection of the tools? Is it the governance? Is it just a, you know, internally socializing this?

Uh, is it a cultural friction? What's, what's stopping you from going faster or scaling faster, I guess? The friction points for me, I would say the speed with which technology is coming at us.

It, it-- And I don't see that as a friction point, as more, may... I, I mean, yes, you could see it as friction, but I see that more of, an opportunity for us to adapt the way that we try to view the full spectrum of what is before us. And, I always w- go back to the, to the CIO Summit a couple of years ago, the Microsoft CIO Summit, and somebody flashed up on the screen a slide that had a logo for every AI company in Silicon Valley.

And it looked like dots, like literal tiny little dots on this PowerPoint slide because... And I can't imagine what that slo- slide looks like two years later. Probably e- you know, constellation.

I don't know. Like, you can't even see, see. It's probably just all dark because there are so many companies.

My point being that people have a lot of ideas, people at Barracuda, people everywhere, because they're, they're seeing a lot of opportunity. And I want to be able to be the, you know, the facilitator of their hopes and dreams to use AI for whatever they want, but I also need to keep it safe and secure, and I need to keep cost consciousness in mind. So there's a balance that we have to strike.

So I would say it's the speed with which these companies are coming into existence, the offerings they have, making sure that I understand, you know, like, what it is that I may wanna make trade-offs about because I, I don't have a portfolio that I can't maintain. And then ensuring that I'm choosing, you know, that I'm keeping cost consciousness in check. All of those are areas of opportunity and potentially friction.

Hey, you know, one, one last question I had was, about where we are in this journey. Specifically, we heard so much about this would be the year of AI, AI agents, although I suspect next year really is, or th- twenty twenty-five was supposed to be. Now I suspect twenty twenty-six is.

And it-- I keep coming across the reports, and I-- you don't have to comment on this, but there was a Goldman Sachs analysis that despite the massive investment in boardroom hype, AI didn't add much to the overall US GDP last year. So I'm wondering, is we-- are we in a kind of a early stage and about to take off in, in terms of use and in terms of implementation of things like agents? I think so.

I think we're really in our infancy. We don't know what the future... Well, we, we have a sense of what the future holds.

Mm-hmm. But to your point, John- Mm. There is, you know, there is a lot of media hype about what AI is doing for our workforce and our companies.

So if you, if you take what you said and kind of transverse it into the media hype about, you know, like, companies that have, touted layoffs related to AI, and you look at some studies that have recently been released. There are many studies that are rooted in real detail and fact that are showing that there is less than one percent of layoffs from twenty twenty-five that are attributable to AI. Um, sometimes at higher salaries, so that's something for us all to keep in mind, where they thought they could kind of lower their, you know, their employee count based on what they think AI can do today.

And if you take all of that into consideration and you answer your question, John, I believe we still have so much to learn about the way AI will operate in our environment. We, uh-- There are agents upon agents that can do things for us. There are ways that we can create agents and have them be part of our organizational structures, actually, y-you know, that can do things, that can actually provide, you know, contextual answers to questions that we have.

Today, we're not there, and I'm not comfortable with do-- you know, allowing that to happen. But that would be the next level up for us- Mm-hmm ... is to have that be part of our, our environment.

So we're right at the beginning of all of it, and I think we're about to come into a, a new stage of bloom with related, related to AI. You know, that's a really, great way, I think, to end this conversation. Um, you know, it's been, it's been really interesting.

And a-as I said earlier, I think that the thing that sticks out to me is the nuance of all of this. It's not, you know, let's, you know, run forward full speed. It's not let's run away and hide.

It's let's figure out what makes sense. Let's figure out how we can work with people, how we can work with tools, how we can work with technology and make this thing useful to us. And I, I, I think that many of our listeners were probably nodding along as you were saying many of those things because we're all seeing it right now.

We all are under, AI mandates and conflicting AI warnings from everyone around us, and we're all trying to navigate through it. Uh, John, yes, I absolutely think that, this may not be the year of AI agents. Maybe that'll be next year, maybe it'll be the year after, or maybe we just won't notice, and it'll just be how things are done.

I think that's the right answer. So thank you. Yeah, thank you so much for joining us, today.

Uh, it's been, it's been great having you, Siri. Um, w-before we go, I know that some folks are gonna wanna continue this conversation and, contact you. Where can people find you, and what do you have going on right now?

Yeah. Thank you for that. So, a couple things.

Before I let you know where you can find me, I'd just like to close out by saying to all of my CIO compatriots out there, I invite you to consider a, a governance structure that could be of any size. Make sure that you consider one, so that you can have review mechanisms, because that is really going to create some safeguard rails for you and allowing you to feel like you can kinda take off. So, that would be my, my final word.

And then in terms of where you can find me, on LinkedIn, that's a very great place to find me. Please do engage with me there. And if you happen to be in Las Vegas, I will be speaking, on April eighth at the ISSA group meeting.

Outstanding. I might, I might try to, to attend that if I can, if I can sneak in. Great.

Um, I will be difficult to find in the real world, in the next few weeks. The events that I'm going to are client events, so they're not really super public. Uh, so they're a little bit top secret, so I can't tell you what.

Um, however, you can always find me on, on X and on LinkedIn, and I would also suggest that you pay attention to some of the, studies and reports that I'll be putting out in the next couple of months. There's a, there's a big one on robotics TAM, so, total addressable markets. Mm-hmm.

We're doing a, a ten-year study on that, that we're gonna be releasing soon. Uh, there's more stuff about PCs, and we have a forecast coming in the next couple of months for all AI devices, including AI PCs, which is gonna be interesting because there's a concern that, rising prices of certain components might act as a break against, PC sales and AI PC adoption. Uh, and so I won't tell you what, what I think, what my assumptions are, but we'll have some results and some data to show you or share with you, very, very soon.

So stay tuned for that. S-hey, so Olivier, by the way, I'm, I'm really interested in the robotics and the AI device report, so please reach out to me on Slack. Um, so I'm- All right ...

mainly available on, on L-- Thank you. I'm mainly available on LinkedIn, at JohnSwartz, one word, and, I'm on the Techstrong websites. There are very-- There are a lot of them in terms of content.

Um, briefly, the next couple of weeks, I'm gonna be here in the Bay Area, going from show to show. Uh, there's a GTC, the NVIDIA show in Santa Clara. JavaOne is the week after in, in, across-- Literally, we live next to the Oracle campus in Belmont, Redwood Shores.

And then I'm gonna go to RSA in San Francisco, and then I'm gonna spend a lot of time in Las Vegas in April and May. Yep. We've got a bunch of people going to RSAC.

I know about that. I also wanna put a little plug in here for AI Field Day, which is the organiz-- the, the event that I organize. Uh, that is gonna be May thirteenth, through fifteenth.

Um, and we are gonna have a really packed agenda. Uh, very excited to be, discussing AI with a bunch of companies. com to learn a little bit more about that.

And of course, you'll find, the AI Field Day sessions on the Techfield Day YouTube channel, as well as right here on Techstrong AI and the Techstrong TV app. So thank you everyone for listening to this episode of the Utilizing AI podcast. Uh, if you enjoyed it, please do subscribe on YouTube or in your favorite podcast application, and consider maybe giving us a rating or a review.

This podcast was brought to you by the analysts and experts from the Futurum Group, where insights meet AI. ai, the Utilizing AI YouTube channel, or the Tekstrong TV app. Thanks for listening, and we'll catch you next Wednesday.

Hey, everyone. Welcome back here to our Predict 2026... Wow, still kinda getting that to roll off my tongue.

Our Predict 2026 edition. Uh, we hope you've enjoyed some of the sessions that we've presented here today. There's certainly been a wide, array of different viewpoints and different areas and, and what the year may have in store, whether you are in the data world, digital world, semiconductors, what have you.

My next guest is yet another one of the, Futurum advisory analysts. I'm very happy to have my friend Keith Kirkpatrick here with us today. Let's welcome him.

Keith, welcome. Thank you for being here on Predict 2026. Thanks.

Keith, before we- Thanks for having me. jump in... Oh, it's my pleasure.

It's our pleasure. Keith, before we jump in, give people briefly, 30, 60, 30 seconds, 60 seconds, a little bit maybe of your title, your specialty, kind of your story arc. Absolutely.

Uh, so I am Keith Kirkpatrick. I'm a research director here at the Futurum Group. I cover enterprise software and digital workflows.

What does that mean? Well, it means any kind of software that you might use within an enterprise, everything from your Salesforce CRM to perhaps Adobe Creative software. That's what I'm looking at, and I'm not just looking at the technology itself, I'm looking at how do people actually interact with and use that software to become more efficient, to become more productive, and hopefully to, to drive the top and bottom line for the business.

And I... My career started out as a journalist, and I still have sort of that inquisitive mindset as I'm looking at all of these companies and technologies. Fantastic.

So Keith, for today's session here on Predict, as we look, you know, short term, near term, we're looking-- we're gonna look specifically of, how will agentic AI impact the licensing and pricing of SaaS soft-software. And I feel like we really can't have any discussion around agentic AI, especially 2026 in mind, without recognizing that, you know, in many year-- in many ways, 2025 was the year agentic AI kind of burst on the screen. But sometimes, you know, with a nice Christmas, New Year's break, you have a chance to reflect, and you realize for all the buzz and the talk around agentic AI, I would not be surprised if the consensus was, well, so far these agents really haven't lived up to the hype.

Yeah. Not that anything can live up to the hype. The hype was so high.

Yeah. But I'm interested in your take on that. What, you know, before we look ahead- Sure ...

let's look back. You know, Alan, you, you really kind of captured it right there. There was a lot of, of talk about agentic AI, about how 2025 was gonna usher in a completely sort of agentic future where you'd have automation here, automation there, and not just automation, but intelligent automation.

And the problem was that if you really look at what actually was released and what actually was put in production, it really wasn't that much. And there's really a couple reasons for that. One is, of course, the technology was very new.

It only acts upon the data that an organization lets it act upon. So the problem is, if you want your agent to handle all of your customer experience or customer service activity, you have to make sure all of that data is made available, and not only that, you need to make sure that it only uses that information and doesn't just grab information forever. And then, you know, if that happens, then you have these, you know, the scenario of agents kind of going wild and pulling out data that's not relevant and really not doing what it's supposed to be doing.

So I think what we had were some examples of some, you know, good pilot programs, very small scale deployments. But I would not say that 2025 was the year that agent-agentic technology really took off and became mainstream. Certainly not in '25.

Absolutely not. Doesn't mean that 2026... It d-doesn't mean that it won't happen in 2026, of course.

But let's, let's talk specifically around the current utilization, future projections for generative AI and agentic AI. Sure. So if we think about where we are in 2026, early 2026, generative AI has been sort of commercially available for going back now about three, three and a half years, depending on when you wanna start the clock.

We're starting to see that become more mainstream because of the fact that everybody from kids in-- at grade school level to people who are, you know, well advanced in their career have at least played around with, whether it's ChatGPT or any one of the other number of models out there, and they're realizing that is very good for certain specific tasks. For example, summarization. Instead of reading through a 30-page report-I can use generative AI to summarize it for me and pull out, you know, the most relevant parts or the most relevant details.

It's very good for doing things like generating, text based on a specific corpus of data. Uh, so there are certainly very, very good use cases for generative AI. Where I think we're still kind of, in that learning phase is figuring out how to do that in a very sort of guardrailed, controlled manner.

What do I mean by that? Well, it means that it's one thing if you are just screwing around with generative AI to, let's say, pull up all the information on a specific topic. But if you wanna use that in a way where it's actually critical that the information is correct, there are a lot of guardrails that need be, need to be put in place in terms of making sure that the information is correct, and it only grabs, the information from which, it's supposed to and doesn't make up or hallucinate other responses.

We're getting there. It's taking some time, but we are getting there. Now, if you wanna switch the conversation to agentic AI, I think this is where we're still a bit...

You know, I, I, I like to use the, the baseball analogy. We're probably in the bottom of the first inning, something like that, in terms of, you know, how far along we are in the game. The reason for that, again, is because if you look at agentic technology, the big challenge there is, again, making sure it's utilizing the right data, only the right data, and is actually following the steps that are required within a specific process to yield the re-desired results.

You know, if you think about agentic technology, we would love it to be sort of like a superhuman way of doing things, that it's able to ingest information from a million different sources and use a reasoning engine to take the right action. That takes time to get that right and to make sure that, you know, that again, it is following all the steps that are required. I mean, if you think about, in regulated industries, there are certain steps that are, that quite honestly are designed to introduce friction into the process to make sure that all of the boxes are checked.

If you think of agentic technology, in some ways, it can actually skip past that, and that becomes a problem. Where I think we're going from here, I think we're going to see improvements. I think we're gonna eventually see, and I'm sure we'll get into this, we'll start to see agentic technology being used for more than sort of very, very basic tasks like I wanna look up my bank balance and then ask me, you know, do I wanna transfer money?

Do I want to, pay a bill? That sort of thing. I think we're gonna get beyond that, but it is gonna take time.

Excellent. I, I agree with you. And, and you know, just the notion of time itself, Keith, in this AI world we're living in, it seems like time is condensing.

Uh, I know nothing goes faster than the speed of light, but, you know, time, time seems to be accelerating. Um, I, I wanna bring it home to SaaS software a little bit, if it's okay with you. Right?

How, how is agentic AI, and let's assume it does get better and mature, more mature in twenty twenty-six, how does that affect the licing- licensing and pricing of, of SaaS software, you think? And, and, you know, let's take it at a macro level, and we can dig into specifics as we go here. Sure.

Well, if you think of traditionally, when you think of SaaS software, it was very much based upon the number of people in your organization who would use that software. That would be you used to have a ratio of one person, one seat license. You paid a monthly fee for that, and that was well and good in terms of the way that software used to be consumed.

Now, as we're starting to get into this era of incorporating AI, whether we're talking about generative AI or really any I-AI or agentic AI, now we're starting to see a greater number of processes that are actually being automated and really, quite honestly, do not necessarily track back to one individual. So that complicates the matter because let's just say, me as Keith, as a user, I might use, a ridiculous amount of generative AI, compared with another worker who maybe only uses it a couple of times a day. So in that sense, a regular sort of blanket seat license might not be the most efficient way to price it.

On the other hand, you may have a situation where you have processes that are fully automated and are not tracked at all to an individual. In that sense, we're starting to see, you know, vendors price it, on a consumption basis using tokens, saying that for this particular action, that is worth five tokens, and they would be charged every time, you know, the model is, is utilized. And then, of course, the, the third way, and this is something that I've been focusing on, is looking at how AI can be priced on an outcomes basis.

And what do I mean by that? It means, let's say, instead of, a human answering my customer support call, an agent might handle that. And if let's say that agent resolves that interaction with no human interaction or no human input, that would be a resolution or an outcome, and the vendor would be able to charge based on that outcome being, successfully completed.

So we're starting to see some different pricing in the market based on how the technology is working. Absolutely. Keith, next area I wanted to touch in on was it's not just an agentIt's the army of agents, if you will, multi-agent.

And any time you have, you know, you know what you learn in business school, any time you have more than two people, you need management, right? And, and, in, in this case, let's call it orchestration. And so how is this...

You know, and everybody's throwing their hat into the ring to volunteer. Not to volunteer. Volunteer seems you don't get paid.

Everybody's throwing their hat into the ring to get paid to be your agent orchestrator, to be your agent manager, your multi-agent manager. Excuse me. How is that whole kind of...

it's more than a cottage industry. How is that whole environment impacting the use of agents in general, as well as the pricing around them? Sure.

A-Alan, you bring up a really interesting sort of scenario that we're entering into. If you look at how vendors are introducing their own agents. So let's say I'm a Salesforce user.

Salesforce has a whole, well, they call it Agentforce, this whole army, as you say, of different agents. As an enterprise user, yes, I might be using Salesforce, but I also might have a, a ServiceNow license and be using some agents from them. I also might have an IT staff, a development staff that has developed agents.

Well, if you don't wanna have a scenario where all of these agents are acting either in conflict with another or are, you know, basically being repetitive in terms of the tasks that you're doing, you need some way to orchestrate and make sure that each agent accomplishes its task without interfering or, or somehow breaking a process that another agent is doing. So basically, like you said, vendors want to be that, you know, each vendor wants to be that agent orchestrator where they manage not only their agents, but third-party agents and homegrown agents. The challenge, of course, is that everybody wants to do it.

And, you know, vendors, they like to say, we like to play nicely in the sandbox with one another, but in reality, everybody is, is highly competing there. And I think what we're going to start seeing are a couple of things. First, on the pricing.

Again, I think we're gonna start to see a greater pivot to outcome-based pricing where you can say, "I have this process that I want done. " And then, of course, you need an orchestration layer in the background to ascertain, you know, who's, you know, which agent is handling which part of that process, and then, of course, divvying up the revenue piece in terms of, you know, which vendor gets paid which part there. I think that is gonna become the big sort of, challenge for vendors as we move forward.

Uh, you know, ultimately, there are very, very, very few enterprises that only use one vendor to handle all their software needs. So I think this is gonna be something that is, especially important to sort of sort out over time as we do start to see agents pl-pl-uh, proliferate. Fair enough.

Keith, let me bring up another subject. Right? I, I, I have to admit, when I first saw agents, the, the AI agent stuff coming out, I had a hard time with that first generation of agents, you know, the difference between them and APIs, if you will, and, and how they APIs talk to each other and, and interact and so forth.

We've come a bit since then, right? I, I get h- what agents are supposed to be doing and what they do, but so many of them are still sort of, let's call them one-trick ponies. Right?

And they, and they're, they're relatively ephemeral. There's not, there's not a, a good reason for them to be persistent, right? But as we add more autonomy to them, now you start seeing a better reason why they should be persistent, why they are more than just one-trick p-p-ponies, why, why they will have the impact that we all expect them to have, right?

However, that's not free or cheap, right? And that's gonna drive pricing and, you know, and then you have the market come into play. Well, if it's too expensive, I'm not gonna be so quick to do it.

And I... so you gotta find the right price that kind of optimizes the, the evolution of these things here. How, how does this all play out in twenty twenty-six and beyond?

I think we're gonna start to see now a situation where if I have agents that are being used to, say, assist human workers, that may be priced in one way. That might be based on a consumption basis, saying, these agents are using this amount of resource, essentially compute power, you know, through the model, and I could see that those types of agents being priced one way. Agents that are much more autonomous in nature, that are designed to accomplish, you know, routine, repetitive tasks, I can see those being priced using an outcome model.

And then, of course, there's gonna be, you know, the cases where you're gonna say, "I know that I am gonna be using a ton of resource here, and it's more efficient if you buy a seat license," price that whatever that might be to cover sort of that, I don't wanna call it all you can eat, but a greater amount of sort of pre-build, AI resource. So in a sense, what we're looking at here, as we continue to see different types of agents being put into the mix, certainly ones that are using, and this is the key, the ability to reason, to take in data-And then actually take action. I think we're gonna see a greater sort of, mix of different pricing approaches being used, to really make sure that each agent is priced or at least positioned, the price being positioned as being most efficient for the customer.

And then of course, the vendor needs to benefit as well. Fair enough. Good, good stuff there.

Keith, got a la-- final sort of category that I wanted us to present today to our audience and that-- and it's this: as we normalize, if you will, AI pricing, because I, I think we all agree, we, we recognize that a lot of these AI companies, you know, it, it, for every eighty cents that they charge, they're using a dollar's worth or more. Yeah. And, and that's, and you know, and it-- I remember back in my dotcom days, we can't make it up in volume.

You're still losing on every transaction. Right. But let's assume, you know, market forces come into play and, and AI, both agentic and generative for that matter, but AI pricing in general has to become more reasonable, more business-like.

It has to be profitable. Right. It has to generate profits, not just top line.

But as that happens, it's, it's gonna have an impact on vendors, on SaaS vendors. It's gonna have an impact on customers, and you know, for some of us it's cost because we're, we're utilizing these agents and stuff, so we gotta figure the real cost, tokenization- Right ... and so forth, tokens into our system, into our cost model.

And then for those of us who are pushing agents, if you will, right, what's the revenue model going forward? How does a more sane or more market savvy view of the AI pricing impact that? I think it's, a-again, to, to what you mentioned earlier, we're still at a, an area where, you know, the resource is not inexpensive yet.

We might, we might get there over time, but it is going to be the result of a confluence of real-- a real increase in demand. And I mean, and that's only gonna come by when a company out there figures out, wow, by a-- by using agentic AI, let's say, to automate a process, that used to take, you know, three human people four hours each day to do, and they can turn that around in, you know, seconds. Right there, you're going to see a real shift in the way, AI companies are able to monetize AI because they're gonna actually see that ROI, and it will become valuable.

Right now, it's still very much based around that sort of consumption model of, I'm using this amount of compute power to do this, like, small task, and yes, you do get a benefit over time if you, you know, do these, small tasks, and you can eventually sort of see how much time you saved. But until you start to see that exponential benefit, I think that's going to be... " And again, and I sound like I'm, you know, a broken record here, but it does go back to that outcome-based approach that I believe will come into the market because AI will be solving a specific problem.

It won't be a simple one. It'll be a much more complex one that there is no other way you can do it except by using that technology. When we hit that point, we're gonna see a shift in the way that AI and generative AI and agentic AI is monetized.

I love it. Keith, that's all I have. We're about out of time for today's session.

I do wanna mention, though, that we do have a video of you with your specific predictions- Yes ... for twenty twenty-six, and we'll have that over in the, in the Futurum, exhibit. We, we-- So all of our sponsors now have, like, booths, virtual booths, call them exhibits, and we have one set up there for Futurum with a lot of, the extra content that's available: reports, videos, et cetera.

And if you head over there, you can get Keith's specific twenty twenty-six, predictions. Keith, thanks for coming on and being part of Predict 2026. Thanks.

Thanks very much, Alan. We'll have to revisit it. My pleasure.

Same here as well. Keith Kirkpatrick. All righty.

Keith Kirkpatrick, Kirkpatrick, Futurum analyst here as part of our Predict 2026. Stay tuned, we have a lot more on Predict 2026, including the winners of this year's DevOps Stars awards.