[00:00:04] to another episode of TechUnhinged where

[00:00:06] tech gets human. I’m your host Rabia

[00:00:07] Javed and joining me today is Luis

[00:00:09] Suarez who is the CTO at HIG Capital.

[00:00:13] Louis spent nearly 7 years as a computer

[00:00:15] forensics investigator ran IT career and

[00:00:18] built systems at Radar. Today he also

[00:00:21] chairs the cyber and technology

[00:00:22] committee on the board of Goodwill

[00:00:23] Industries of South Florida. Welcome to

[00:00:26] TechUnhnged Luis.

[00:00:27] >> Hello. Hi. Well, uh, we’ll start off

[00:00:29] with the icebreers. Louie, you have

[00:00:31] worked two full-time jobs for a good

[00:00:33] seven years straight. As a forensic

[00:00:35] investigator and consulting director,

[00:00:38] both at the same time. How did it keep

[00:00:40] you sane? You know, that when when when

[00:00:42] I was doing that, one kind of really led

[00:00:44] in into the other, which was great. the

[00:00:46] the forensic side of it was the fun, the

[00:00:50] spontaneous, the uh whether it was for

[00:00:52] the SEC, for the IRS, for civil

[00:00:54] litigation, and those were in the early

[00:00:56] days of computer forensics and and that

[00:00:58] was really testing a lot of the

[00:00:59] technical skills that that was emerging

[00:01:02] in those days. So, I I really enjoyed

[00:01:04] it. That was assignment based, which was

[00:01:06] great. And then there was the the day

[00:01:08] work, which was really supporting Latin

[00:01:10] American headquarters uh in the South

[00:01:12] Florida area for a long time and helping

[00:01:13] them out. And it it just it was really a

[00:01:15] fun time where the technology was not

[00:01:17] where it is today. So a lot of it was

[00:01:19] really emerging at that point.

[00:01:20] >> No, I think that’s great. Also, you

[00:01:22] know, an interesting another icebreaker.

[00:01:24] You took Harvard’s program on

[00:01:26] negotiation and competitive decision-

[00:01:28] making. So, out of all the leadership

[00:01:31] skills you could have invested in, why

[00:01:33] did you pick negotiation particularly?

[00:01:35] >> You know, um it’s it’s probably the best

[00:01:37] course I’ve ever taken because what I

[00:01:40] really liked about it is our day-to-day

[00:01:42] world is about negotiation. whether you

[00:01:44] realize it or not, whether it’s in in

[00:01:46] bridge building, whether it’s dealing

[00:01:48] with the other CXOs at HIG, or it’s

[00:01:51] dealing with with the clients, internal

[00:01:53] clients, and it’s it’s always about

[00:01:55] really learning how to come to an

[00:01:57] agreement where both sides feel as if

[00:02:00] that they got something. And it really

[00:02:02] has has worked well. One of the things I

[00:02:03] really enjoyed in the course was the

[00:02:06] concept of that when you’re negotiating,

[00:02:07] which is just like in a relationship,

[00:02:09] you’re going to come back to that person

[00:02:11] again. So if if you really um have a bad

[00:02:13] negotiation or a bad relationship

[00:02:16] conversation, it’s not a oneanddone. So

[00:02:18] you have to build on it. And and that

[00:02:19] course was really fantastic about that.

[00:02:21] And and I use pieces of that course in

[00:02:23] my daily life, whether it is actual

[00:02:25] negotiation with vendors or it is just

[00:02:28] in the day-to-day negotiation of being a

[00:02:31] a CTO. You know, we we don’t run this uh

[00:02:33] in in one direction. All conversations,

[00:02:36] all all uh decisions are birectional. So

[00:02:39] those skill sets were just immensely uh

[00:02:42] useful for me.

[00:02:43] >> Moving a bit more into the context

[00:02:44] setting of the topic that we have in

[00:02:46] today. So Louie IDC’s 2024 survey found

[00:02:49] out that only 20% of the CIOS rank tech

[00:02:53] debt as a top priority. What is it about

[00:02:57] AI specifically that is forcing the

[00:02:59] boards to start thinking about it

[00:03:01] finally?

[00:03:02] >> Yeah. So, you know, if AI is going to do

[00:03:04] one thing and one thing very quickly, it

[00:03:06] is going to highlight your tech debt. As

[00:03:08] you go down the AI journey, your tech

[00:03:10] debt, whether it’s defined as systems

[00:03:13] that are not ready or data that’s not

[00:03:16] ready, AI will really magnify and

[00:03:19] highlight that right out of the door.

[00:03:21] You know, it really will point it right

[00:03:23] away. And I think the organizations that

[00:03:24] are being successful as they’re

[00:03:26] embarking on this AI journey, they are

[00:03:28] examining their tech debt and trying to

[00:03:30] modernize it, trying to decommission it,

[00:03:33] trying to remediate it as best they can.

[00:03:35] Otherwise, the AI initiatives will not

[00:03:37] be as successful. We have some systems

[00:03:39] that have been around for quite a bit of

[00:03:41] time and those systems have proved

[00:03:43] challenging to really integrate with AI.

[00:03:46] Our more modern systems, we’ve been able

[00:03:47] to plug them in via MCP connectors

[00:03:50] relatively quickly and get moving. The

[00:03:52] older ones have been a challenge. We’ve

[00:03:54] had to get creative. We’ve had to

[00:03:56] upgrade some of those systems. Probably

[00:03:58] I I would say the the tech debt related

[00:04:00] systems probably take four to five times

[00:04:03] as much effort as as our modern systems.

[00:04:06] Uh they they just take longer. They’re

[00:04:08] not ready.

[00:04:08] >> Yeah. Yeah. Also, another interesting

[00:04:10] questions. Uh Louie, you’ve watched

[00:04:12] enterprise tech move through 2008,

[00:04:16] mobile, cloud, and SAS. What shortcuts

[00:04:19] did companies get away with that they no

[00:04:23] longer can get out of with AI? Now,

[00:04:25] >> the big thing that’s happening with

[00:04:26] Genai now, it is moving so fast. There

[00:04:29] are no shortcuts. You have to really

[00:04:31] plan and and it’s really about the the

[00:04:33] age-old PPT, people, process, and

[00:04:36] technology. And the people part, the

[00:04:38] culture part of it more than ever, you

[00:04:40] have to invest in the if the firm is not

[00:04:43] ready from a cultural perspective to

[00:04:44] embrace AI, it will just struggle to it.

[00:04:47] You know the the shortcuts are really

[00:04:49] about investing upfront in the

[00:04:51] communication in getting the people

[00:04:54] invested from top down. You really have

[00:04:56] to make them part of the process

[00:04:58] otherwise the resistance will come in

[00:05:00] from every facet to it. It’s just moving

[00:05:02] so fast and and some and some people are

[00:05:04] threatened by it. And so it’s the

[00:05:05] community more than ever this technology

[00:05:07] is about communication and getting

[00:05:09] people to be part of this rapidly moving

[00:05:12] evolution. So you know the experts say

[00:05:15] that CTO used to be the person who kept

[00:05:18] the systems running in a company you

[00:05:20] know when it came to it. Now they are

[00:05:23] the person deciding where the company’s

[00:05:25] next big investment is going to land.

[00:05:28] First, do you agree with this? And

[00:05:30] second, what forced that particular

[00:05:33] shift in your view?

[00:05:35] >> Listen, absolutely. Absolutely. Days of

[00:05:37] CTO, CIOS being maintenance folks,

[00:05:40] keeping the the lights on are long gone.

[00:05:43] AI technology now can and should be a

[00:05:46] competitive advantage for the business.

[00:05:49] And that means being at the table having

[00:05:51] a discussion in some cases leading the

[00:05:54] business towards te technological

[00:05:56] evolution that will give the business

[00:05:58] whether it’s a financial advantage in

[00:06:00] terms of efficiencies and in some cases

[00:06:02] again we’re a private equity firm. We

[00:06:04] we’re invested in over a 100 portfolio

[00:06:06] companies. We are in some cases building

[00:06:09] new products thanks to the technology

[00:06:11] thanks to the AI. You know the AI is

[00:06:13] opening doors that that were just cost

[00:06:15] prohibitive before. And so the CTO’s,

[00:06:17] the CIOS that are going to be successful

[00:06:19] are the ones that are true business

[00:06:22] partners, true business leaders in the

[00:06:24] firm more than ever before.

[00:06:26] >> And you know, Louie, for our audience,

[00:06:28] you know, since we are discussing um

[00:06:30] tech debt in particular, for the people

[00:06:32] who still are probably unable to

[00:06:34] understand what that means or maybe it’s

[00:06:36] a bit of a fancy term for them, if you

[00:06:39] as a CTO would have to explain it to the

[00:06:41] layman.

[00:06:42] >> Sure. you know, tech that I I usually

[00:06:44] use the frame of reference and you can

[00:06:46] pick a different amount of time frame.

[00:06:48] But if you if you’ve had a system for X

[00:06:50] number of years, I put that on my yellow

[00:06:52] spotlight saying this is probably going

[00:06:54] to qualify as tech debt. If you haven’t

[00:06:56] innovated it, if you haven’t upgraded,

[00:06:58] if you haven’t done anything with it in

[00:07:00] say 4 years, pick a number that’s

[00:07:03] probably right, that puts it in that

[00:07:05] quadrant there of this is probably tech

[00:07:07] debt. And usually the the tech that is

[00:07:09] those systems that you cannot integrate

[00:07:12] with easily. Nowadays no system lives by

[00:07:15] itself. Those days of having these

[00:07:16] systems that are islands are gone. We

[00:07:18] have to because of data because of AI

[00:07:21] because of the new technology that

[00:07:23] facilitates these systems are all an

[00:07:25] ecosystem versus a set of islands that

[00:07:28] are somewhat independent. It is now an

[00:07:30] interconnected ecosystem that data moves

[00:07:32] within systems and the data becomes the

[00:07:35] currency for AI. And the moment you have

[00:07:37] a system that if you look at it and it

[00:07:39] is an isolated system that nothing else

[00:07:42] talks to, that’s tech debt. I love using

[00:07:44] that that island analogy because it

[00:07:46] really it’s about taking technical terms

[00:07:48] and making them really techneutral in

[00:07:50] the conversation. So whether I’m

[00:07:52] speaking to a CEO who is very

[00:07:55] technically literate or or is not,

[00:07:57] they’re going to understand it. And at

[00:07:59] that point, we we are talking the same

[00:08:01] language and and at that point we’re

[00:08:03] looking at the same goals on how we get

[00:08:05] there. When is there a point Louie when

[00:08:07] a founder or a company owner could

[00:08:11] actualize that they are under deep tech

[00:08:14] debt and AI is not the solution right

[00:08:17] away because probably there are steps

[00:08:20] for for them to move towards AI too

[00:08:22] because probably first step is

[00:08:24] modernization perhaps. How how would you

[00:08:26] comment on that? the the moment the

[00:08:28] technology starts to become a limiting

[00:08:30] factor when you’re talking about here

[00:08:32] are the business goals and you start

[00:08:34] seeing well this system can’t do this or

[00:08:37] we don’t have the ability to do this.

[00:08:38] The moment you start having the nos that

[00:08:41] is the clearest sign that we are in need

[00:08:43] of modernization whether it is

[00:08:45] modernization of a platform or a

[00:08:48] complete enterprisewide modernization

[00:08:50] that discussion needs to start

[00:08:51] immediately. that journey needs to start

[00:08:53] because the longer we delay that journey

[00:08:56] the longer it is before we can start

[00:08:58] talking about modern technology whether

[00:09:00] it’s AI whether it’s a bet better data

[00:09:03] stack whether it’s automation you know

[00:09:05] I’m I’m always fearful when we start

[00:09:07] talking about AI for the sake of AI you

[00:09:09] know you don’t want to be a hammer

[00:09:10] looking for a nail the the goal should

[00:09:12] be we need to provide value to the

[00:09:15] business we need to be a differentiator

[00:09:18] for the company and the moment the

[00:09:19] technology is not able to do that that

[00:09:22] tech stack that platform is tech debt.

[00:09:24] It’s a liability and I think that that’s

[00:09:26] the way to really define it. When you

[00:09:27] have a system or systems that are now a

[00:09:29] liability for the goals of the business,

[00:09:32] that is tech debt in it in its widest

[00:09:34] sense.

[00:09:35] >> Absolutely. And Lou, you spent nearly 7

[00:09:37] years at Forensic Technology Group as a

[00:09:40] computer forensics investigator. First

[00:09:42] of all, I’d like you to explain it for

[00:09:44] our audience. What was it like? You

[00:09:45] know, how is it different working in

[00:09:48] forensics as a computer investigator? So

[00:09:51] you know computer forensics will

[00:09:53] probably the precursor to a lot of what

[00:09:55] we deal today with cyber security which

[00:09:57] is when you have an event and whenever

[00:09:59] you bring a forensics person involved

[00:10:01] you’ve had some sort of event cyber

[00:10:02] event operational event technology event

[00:10:05] and it was really playing detective. It

[00:10:07] was using technology to go ahead and put

[00:10:09] together what happened to put together a

[00:10:12] set of data that supports what the legal

[00:10:16] or civil investigator was was looking

[00:10:19] for. And it literally what was

[00:10:21] fascinating about it is that you never

[00:10:22] really knew what you were walking into

[00:10:24] until you got there. You you could be

[00:10:26] talking about looking at a couple of

[00:10:28] laptops and a server. It could be at at

[00:10:30] its worst we had to deal I had to deal

[00:10:32] with 100 laptops and 50 servers to go

[00:10:35] ahead and analyze them and put a story

[00:10:38] together that the data supported of what

[00:10:40] happened, what the criminals did or what

[00:10:42] happened from a civil litigation

[00:10:43] perspective. And those skills ended up

[00:10:45] becoming important in my career because

[00:10:47] when you’re working in technology,

[00:10:48] you’re always having to piece together

[00:10:50] what you want to do or what happened

[00:10:52] using data as your information source.

[00:10:55] >> AI is replacing some of the engineers

[00:10:57] who used to do the unglamorous work of

[00:11:00] paying down tech debt. Are we automating

[00:11:02] away the people who can actually fix it?

[00:11:04] >> No. I you know, so I think we are having

[00:11:07] a shifting of resources. I think you

[00:11:09] have some roles that the that will

[00:11:11] probably be eliminated. I agree with

[00:11:13] that. We we we can’t run away from that.

[00:11:15] But I think we’re also opening up

[00:11:17] opportunities for other roles to really

[00:11:19] expand what they’re doing and shift

[00:11:21] people. You know, one of the big things

[00:11:22] that that’s been talked about is that AI

[00:11:24] will replace development and that all

[00:11:26] development will be AI based. I think

[00:11:28] we’re just going to have different kinds

[00:11:29] of development. I think we’re still

[00:11:31] going to have developers. I think

[00:11:32] developers now are going to have more

[00:11:34] tools at their disposal than ever

[00:11:36] before. I think we’re going into an age

[00:11:38] of just massive creativity that’s being

[00:11:40] unlocked with these tools. And I also

[00:11:42] think it’s going to facilitate getting

[00:11:44] out of tech debt. If you have a system

[00:11:46] that was in a language that was 20 years

[00:11:49] old and you need to do a conversion

[00:11:51] project and cost was a problem, AI will

[00:11:53] now unlock that and allow you to take an

[00:11:56] 18-month development uh code conversion

[00:11:58] project and have that happen in 90 days.

[00:12:01] And that is an opportunity that is not a

[00:12:03] crisis of labor that allows us to move

[00:12:06] our labor from being conversion to say,

[00:12:08] okay, we’ve now converted this. Now what

[00:12:10] can we do? we’re now opening up huge

[00:12:12] opportunities.

[00:12:13] >> Now we’ll dive a bit into um how

[00:12:15] innovation is meeting financial

[00:12:18] discipline. So one of the researchers

[00:12:20] found that tech debt consumes around 20

[00:12:22] to 40% of enterprise development time.

[00:12:25] If that’s true and if you would agree to

[00:12:27] that, do you think that fixing it

[00:12:29] doesn’t slow AI down but it gives you

[00:12:31] the time back? And the next year to this

[00:12:34] question is that why is the speed versus

[00:12:36] foundations tradeoff still being framed

[00:12:39] as a real choice for the CTO’s.

[00:12:41] >> So to to answer the first question, you

[00:12:43] know, the the tech that needs to be

[00:12:46] resolved needs to be innovated upon and

[00:12:48] and it is a blocker and and so we we

[00:12:50] need to move past it to go ahead. I I I

[00:12:52] I think you know if if we’re not doing

[00:12:54] this then we are limiting ourselves in

[00:12:57] what we can do. And to your second

[00:12:59] point, you know it’s the age old

[00:13:00] question, right? It’s it’s speed versus

[00:13:03] thoroughess. It’s speed versus making

[00:13:05] the right decisions. And I I think those

[00:13:07] are not those are not contradictory.

[00:13:09] Those go hand inand that that is all

[00:13:11] about proper planning, strategic

[00:13:13] planning. If you’re going to go down

[00:13:15] this road, not if when you must of

[00:13:18] moving away from your tech debt, you

[00:13:20] need to take that extra pause and figure

[00:13:22] out what is this new foundation that you

[00:13:25] are designing to it. you don’t want to

[00:13:27] happen is that you lock yourself into

[00:13:30] another tech debt round 2 3 4 years

[00:13:32] later and so taking that extra time to

[00:13:34] think about it to say you know and and

[00:13:36] it’s not just doing it for the sake of

[00:13:38] AI it’s doing it for proper architecture

[00:13:40] proper future modernization and so

[00:13:43] making tactical decisions without a

[00:13:45] strategy is what leads to more to

[00:13:47] another round of tech debt at a future

[00:13:49] so I I I think there you can move you

[00:13:51] can move efficiently but moving super

[00:13:53] fast without that strategic conversation

[00:13:56] put you in a position that you will be

[00:13:57] having the same conversation at a later

[00:13:59] date. And the way the cycles are moving

[00:14:01] quicker and quicker, it won’t be 5 years

[00:14:02] from now. It could be 2 years from now

[00:14:04] because we’re no longer in these cycles

[00:14:06] of putting in a system for 10 years and

[00:14:08] 5 years. You know, AI has really

[00:14:10] compressed those timelines dramatically.

[00:14:13] >> So being the CTO yourself at this

[00:14:14] position, how adaptive were you to this

[00:14:17] particular technology and now at your

[00:14:19] particular role, if you can talk about

[00:14:21] it, how have you implemented it at an

[00:14:24] org level?

[00:14:25] >> Yeah. So, you know, the the the first

[00:14:27] thing was, you know, it’s okay to be

[00:14:29] uncomfortable as a CTO, CIO. We have to

[00:14:32] embrace if you’re being uncomfortable

[00:14:33] with new technology, that means it’s

[00:14:35] probably a disruptor and you need to

[00:14:37] identify that. And we early on

[00:14:39] identified when when the big uh chat GBT

[00:14:42] moment came out with Genai. We saw like

[00:14:45] a lot of my peers, this is interesting.

[00:14:47] This is doing something. So, we started

[00:14:49] experimenting with it right away. And I

[00:14:51] think that was very important. And we

[00:14:53] went through the cycle of building our

[00:14:54] own then realizing in our case that we

[00:14:56] wanted we’d rather buy versus build on

[00:14:58] the Genai technology. And once we

[00:15:00] understood the potential ramification we

[00:15:03] started negotiate not negotiating we

[00:15:05] started discussing with our executive

[00:15:08] team about this technology about what it

[00:15:10] was going to do. We were educating them.

[00:15:12] That goes to what I was talking about

[00:15:13] which is a cultural problem. You have to

[00:15:15] educate. You have to bring them along

[00:15:17] for the journey. And you need to do it

[00:15:18] in a way where you’re not just going to

[00:15:20] go from one day to the next say, “Oh,

[00:15:22] we’re going to implement this AI.” We

[00:15:23] needed top- down mandate. We needed our

[00:15:26] CEO to say, “We will be an AI first

[00:15:30] company. We will be an AI leaning

[00:15:31] company.” And and we’ve had those

[00:15:33] statements from our CEO. We’ve also had

[00:15:35] the firm invest in the people. I know we

[00:15:38] keep talking about AI, we keep talking

[00:15:39] about technology, but it is about

[00:15:41] people. And so, investing in training,

[00:15:43] and we’re not talking about investing in

[00:15:44] a 20-minute video that we make everyone

[00:15:46] watch. We are talking about investing in

[00:15:48] 8 to 10 hours a year per employee and we

[00:15:51] don’t know if that’s the right answer

[00:15:52] but that’s where we’re starting of

[00:15:54] differing levels of how to use this

[00:15:56] technology how to get comfortable with

[00:15:58] it how to embrace it that is is is where

[00:16:01] the differentiator is and that’s where I

[00:16:03] think where we’ve been successful and we

[00:16:04] have been injecting AI in every facet of

[00:16:07] our work streams what we we recently

[00:16:09] there was a press release that was put

[00:16:11] out we invested in a product called TL

[00:16:13] IQ which is for our due diligence

[00:16:15] investigation And that has been a

[00:16:18] dramatic enhancer for us because we want

[00:16:21] to look at more deals. We want to

[00:16:22] process more data. And so this is a

[00:16:24] tool. This is not replacing our animal.

[00:16:26] This is a digital assistant. This allows

[00:16:28] them to take and leverage AI across

[00:16:30] gigabytes and gigabytes of data that we

[00:16:33] receive for any prospective deals. We’re

[00:16:35] now able to drive quicker insights,

[00:16:37] better insights. We’re able to answer

[00:16:39] questions that the investment committee

[00:16:41] may come. We’re also able to run down

[00:16:43] different thesis that we come up with on

[00:16:45] on these businesses and it’s really been

[00:16:47] transformative and we’re only at the

[00:16:49] beginning stages of that. We we think

[00:16:51] there is more to come with this.

[00:16:52] >> So Lou, do you think that tech debt is

[00:16:54] only just an enterprise level problem?

[00:16:57] >> No, tech debt is an organizational

[00:17:00] problem in pockets and and I I think the

[00:17:03] the other facet of tech debt that we

[00:17:04] haven’t spoken about is just using

[00:17:06] applications in a way that they’re not

[00:17:09] designed to be used anymore. One of the

[00:17:11] biggest tech debts that any financial

[00:17:13] services firm has is that old tool

[00:17:15] called Excel. We all love to use Excel

[00:17:17] for things that it’s really beyond what

[00:17:18] it was initially for. And so that really

[00:17:20] lends itself into that people process

[00:17:22] and technology. You have to redefine the

[00:17:25] process, educate the people, and move

[00:17:27] them to a different technology. And

[00:17:28] while Excel is a great tool, it’s not

[00:17:30] going away. Excel is not a database.

[00:17:32] Excel is not a workflow tool. And I can

[00:17:35] go on and on. And that’s where we need

[00:17:37] to really educate the folks and say that

[00:17:39] was a great idea for the last five years

[00:17:41] on Excel but now we we need to modernize

[00:17:43] we need to move away from that here’s

[00:17:45] the right way and that comes into

[00:17:46] educating training upskilling the people

[00:17:48] giving them something new without it

[00:17:50] putting it as as you said earlier the

[00:17:52] shiny new toy you know the shiny new toy

[00:17:54] becomes dull after a while so it needs

[00:17:56] to be a sustainable process and

[00:17:58] technology that they see the value and

[00:18:00] if they don’t see the value they’re not

[00:18:01] going to adopt it

[00:18:02] >> prior to AI there was still tech debt

[00:18:05] out there right it because it was

[00:18:06] probably everywhere but then people or

[00:18:08] people or owners or CTOs or CIOS were

[00:18:11] they just ignorant towards it

[00:18:13] >> you know so first of all you know tech

[00:18:15] debt is is a comb the ramification of

[00:18:18] tech debt is a combination of the tech

[00:18:19] debt and as I’ve been saying the people

[00:18:21] the culture the process to it and

[00:18:23] they’re all tied together so I I think

[00:18:25] as as you go down the tech debt journey

[00:18:27] we need to break some of the old

[00:18:28] processes rebuild them and that in some

[00:18:31] cases allows you to free yourself at

[00:18:33] tech debt from it I don’t think CTO’s

[00:18:35] were ignoring in tech debt. I think

[00:18:37] because prior to to Gen AI and the speed

[00:18:39] that it’s moving, you had the luxury of

[00:18:41] time, you had the advantage of of

[00:18:43] saying, well, this system is going to be

[00:18:45] fit for purpose for another 5 years. As

[00:18:47] we said earlier in this conversation, AI

[00:18:49] has changed all of that. AI is now

[00:18:51] exposing no, this tech debt, this

[00:18:53] platform, this system, this process that

[00:18:55] we’ve wrapped around this tech is no

[00:18:57] longer viable if we want to modernize.

[00:19:00] And and as I said, you know, the comment

[00:19:01] I’ll say and I I’ll say it again is that

[00:19:03] AI is a very good is very good at

[00:19:05] putting a spotlight on tech debt and and

[00:19:08] data inefficiencies and broken

[00:19:09] processes. And so I think what is

[00:19:11] happening is that it is taking away from

[00:19:13] the CIOS and CTO’s the luxury of time.

[00:19:16] We no longer have as much time. We’ve

[00:19:18] gone from dealing in projects that are

[00:19:20] in years to dealing now in projects that

[00:19:22] are in months and sometimes it feels

[00:19:24] like projects that are in days with it.

[00:19:26] >> Yeah, absolutely. But how how does one

[00:19:28] really um stop themselves from getting

[00:19:32] themselves lost within that rabbit hole

[00:19:34] of AI?

[00:19:35] >> That comes down to leadership style and

[00:19:38] the people you surround yourself with.

[00:19:39] So I’m very fortunate to have a very

[00:19:42] good team that has been on this journey

[00:19:44] with me for quite a bit of time and you

[00:19:46] know I I lean into them. You know, I was

[00:19:48] I was taught a long time ago by a mentor

[00:19:51] of mine that you should always hire

[00:19:52] people that may be smarter than you,

[00:19:54] that may know more about a particular

[00:19:56] technology. And so, I’ve really made it

[00:19:58] a point when we recruit to hire smart,

[00:20:01] intelligent, hungry people in there. And

[00:20:03] and that is my nucleus of when we’re

[00:20:06] dealing with these technologies, it is

[00:20:08] not the CTO making a decision. It is the

[00:20:10] CTO exploring the solution with my team,

[00:20:13] whether it’s my cyber team, my

[00:20:15] infrastructure team, my operations.

[00:20:16] Sometimes it’s with all of them in a

[00:20:18] room walking through the problem because

[00:20:20] again as I said earlier we now live in

[00:20:22] the world of ecosystems and making one

[00:20:24] change is no longer an isolated event.

[00:20:27] So we need to have those conversations

[00:20:29] and in those conversations the solution

[00:20:31] will evolve and so it’s always

[00:20:33] interesting to see what you walk in the

[00:20:35] you see a diagram when you walk in the

[00:20:36] room of what you’re thinking you’re

[00:20:38] going to do and after you have a good

[00:20:40] deep honest strategy session you look at

[00:20:42] the final picture and it’s changed it’s

[00:20:44] evolved and you actually have a better

[00:20:46] picture and it’s not the CTO’s picture

[00:20:48] it is the technology team’s picture the

[00:20:50] technology team’s vision everyone has

[00:20:52] bought into it has put something into it

[00:20:54] and everyone owns a piece of it and

[00:20:56] that’s how you get to success and that’s

[00:20:57] also how you’re kept honest.

[00:20:59] >> If you had to comment, they say that

[00:21:00] tech debt is the new technical risk on

[00:21:03] the board agenda. Would you agree

[00:21:06] strongly agree or is that

[00:21:08] oversimplifying the shift?

[00:21:09] >> I don’t think it’s oversimplifying it. I

[00:21:11] I I think you need a concept that

[00:21:13] everyone can relate to can it resonates

[00:21:16] and so I I think if you over complicate

[00:21:18] it, it becomes this long spaghetti of of

[00:21:22] technical jargon that really does not

[00:21:24] allow a board to say, “Oh, okay. I

[00:21:26] understand why why this is I I think if

[00:21:28] you keep it to terms that that are

[00:21:30] relatable, right? Debt is not always a

[00:21:33] good term. When you say tech debt, you

[00:21:35] get their attention right away say,

[00:21:36] “Okay, this is something we need to

[00:21:37] resolve.” And then once you have their

[00:21:39] attention, you can make the board again

[00:21:41] part of the journey. I I’m very big

[00:21:43] about as as I’ve said here, these

[00:21:44] collaborative approaches where it’s not

[00:21:46] the journey of one CXO, it’s the journey

[00:21:49] of the team, whe whether it’s the board,

[00:21:51] whether it’s the leadership of of the

[00:21:54] firm on this. So at that point it is not

[00:21:56] a technology solution, it’s not a CEO

[00:21:59] solution, it’s a company solution. And

[00:22:01] when you get everyone involved, these

[00:22:03] projects don’t don’t go perfect. And so

[00:22:05] when you build that collateral upfront

[00:22:08] of getting the buy in when there is a

[00:22:10] problem, they’re part of the solution to

[00:22:11] help you. Whether it’s you need a little

[00:22:13] bit more funding or you need more time

[00:22:15] or the solution requirements have

[00:22:17] changed because the business doesn’t

[00:22:18] stop changing while you’re moving the

[00:22:20] tech debt or you’re implementing a new

[00:22:22] solution. So you need those partners all

[00:22:24] along the way. It’s that collaboration.

[00:22:26] >> So you know we did talk at length about

[00:22:28] the bright side of AI right but let’s

[00:22:32] talk about where it hallucinates and

[00:22:34] then how most of the people are unable

[00:22:36] to distinguish between that. So as a

[00:22:39] leader within your practice how would

[00:22:41] you comment on AI hallucinations and how

[00:22:44] do you get a control of it?

[00:22:46] >> So the first point is hallucination is

[00:22:48] real. It’s not fiction. It happens even

[00:22:51] when you have a very curated set of

[00:22:53] data. Your prompting can create

[00:22:55] hallucinations and we’ve seen it. So the

[00:22:57] first way we tackle this is education.

[00:22:59] We educate as it’s part of the training

[00:23:01] that that I mentioned. We make it very

[00:23:03] clear to people. We put it on on their

[00:23:05] radar. This can happen that that you

[00:23:07] need to check your work that comes out

[00:23:09] of AI. It is a great tool. It is not

[00:23:12] foolproof. It gets better every day, but

[00:23:14] there’s still the degree of that. So it

[00:23:16] really is the education of it uh and the

[00:23:18] reinforcement and and something as

[00:23:20] simple as we put a disclaimer at the

[00:23:22] bottom that it says please check your

[00:23:23] work. There is a possibility of

[00:23:25] hallucination or or something like that.

[00:23:27] I forget the exact wording. And so the

[00:23:28] idea is you put it there but you have to

[00:23:30] be careful. You can’t make it to the

[00:23:31] point that then people do not trust the

[00:23:33] technology. So it is a fine line but it

[00:23:34] it really is the people part of it. They

[00:23:36] need to be aware that like all tools you

[00:23:39] know all tools can can make mistakes and

[00:23:41] so you just need to make them aware of

[00:23:43] it.

[00:23:44] >> Yeah. So, did you happen to have any of

[00:23:46] your own personal AI hallucination

[00:23:48] moment that you can share?

[00:23:49] >> Oh, sure. I I’ve had where um I do some

[00:23:52] of my my vendor research before I walk

[00:23:54] into any meeting. I use my tools as my

[00:23:56] digital assistant and I will ask for

[00:23:59] give me a summary of of this product.

[00:24:01] How would it apply? There are times that

[00:24:03] it will hallucinate and even though I’ve

[00:24:05] told it this is what my tech stack is

[00:24:07] and it will infer that I have other

[00:24:09] things in my text stack. So, I will have

[00:24:11] to go back and ask it, why did you do

[00:24:12] this or why did you do that? And I’ve

[00:24:15] seen it come back and say, “Oh, you’re

[00:24:16] right. I shouldn’t have said that.” It

[00:24:18] happens more often than people realize,

[00:24:20] but I I think you need to be aware of

[00:24:21] it. You need to challenge when you’re

[00:24:23] using the Genai tools on what it gives

[00:24:25] you. And I think it just makes you a

[00:24:27] better user of the tools. It is I I

[00:24:29] think we are in the age of the digital

[00:24:32] assistant. I think for some specific

[00:24:34] uses you can say okay you’re going to

[00:24:37] run this and the answers are to be

[00:24:39] trusted because you give it a lot of

[00:24:41] guardrails limited set of data and

[00:24:43] that’s what that’s what AI does really

[00:24:44] well but we’re using the Gen AI tools

[00:24:46] and it’s reasoning across internet data

[00:24:48] it’s reasoning about large amounts of

[00:24:50] data just like if someone were to give

[00:24:52] you what I tell folks is imagine that an

[00:24:54] intern gave you this report or a junior

[00:24:56] analyst gave you the report you’re going

[00:24:57] to look at it with a critical eye and

[00:24:59] you should not give up that critical eye

[00:25:01] you should look at it and Just like you

[00:25:03] would question someone, question the

[00:25:04] output. Validate it. You’re still going

[00:25:06] to save time. You’re still going to save

[00:25:07] vast amounts of time. You’re still going

[00:25:09] to get a fantastic work product in the

[00:25:11] end. But do your due diligence. You

[00:25:12] don’t want to be that person that turns

[00:25:13] in your analysis to a superior and they

[00:25:16] find an error within 2 seconds. Kind of

[00:25:18] embarrassing. When you look at high

[00:25:20] performing teams, when you look at

[00:25:21] successful CIO, CTO’s, CISOs, chief data

[00:25:25] officers, you know, all the all the

[00:25:26] technical roles, leadership is about

[00:25:28] really developing your team and your

[00:25:30] talent. And there’s been over the

[00:25:32] decades different things that we’ve

[00:25:33] focused on, but at the end of the day,

[00:25:35] it’s developing the people. And to your

[00:25:36] point, yes, a new generation comes in. A

[00:25:38] new generation comes in under every one

[00:25:40] of our watches. And we just need to take

[00:25:43] what they do really well and then

[00:25:44] transfer what the previous generation

[00:25:46] did really well and make them better.

[00:25:48] That that is our responsibility as

[00:25:50] leaders. While we look forward and see

[00:25:52] what wins the next decade in tech, we

[00:25:56] see that AI is automating decisions

[00:25:58] humans used to make pretty rather slowly

[00:26:00] at times. What does that mean for how

[00:26:02] leadership teams actually monitor price

[00:26:06] and report from here on?

[00:26:08] >> You know, you’re asking me 5 years, 10

[00:26:10] years from now, that that is such, you

[00:26:12] know, a long window now because of the

[00:26:14] way that that technology is moving. I

[00:26:16] mean, we’ve gone over the last two and a

[00:26:18] half years from innovation that comes

[00:26:20] out once a year or moves on a yearly

[00:26:22] basis to right now it’s moving on a

[00:26:24] weekly basis. I think you know how we

[00:26:26] price in is flexibility, adaptability,

[00:26:29] really keeping an eye towards innovation

[00:26:32] more so than ever before. I think you

[00:26:34] know the the next generational group is

[00:26:37] really going to be about ultimate

[00:26:38] flexibility and adaptability. I think as

[00:26:41] we move our next generation into these

[00:26:43] careers, it’s the people that love

[00:26:45] change and and change like we’ve never

[00:26:47] seen before. People that really want to

[00:26:49] work with the next tool and embrace it.

[00:26:52] Uh not because it’s a shiny tool, but

[00:26:54] because they want to do more with it. I

[00:26:56] think this next crop of incoming CIO,

[00:27:00] CTO’s are going to have the most

[00:27:01] versatile, creative, and powerful

[00:27:03] workforce that we’ve ever had because of

[00:27:05] these tools. We just need to get this

[00:27:07] workforce ready. as you just mentioned,

[00:27:08] you know, they they overly reliant on it

[00:27:10] and we can’t lose sight of the

[00:27:12] traditional skill sets that really are

[00:27:15] complimentary to this. They’re they’re

[00:27:16] they’re not uh diametrically opposed.

[00:27:18] They’re not those pieces are not being

[00:27:20] replaced by AI. I think they’re being

[00:27:21] enhanced and we just need to as leaders

[00:27:23] guide that next generation to really

[00:27:25] move into that.

[00:27:26] >> Absolutely. And you know, if you had to

[00:27:28] pick Louisie, what will separate the

[00:27:30] survivors from the winners in enterprise

[00:27:32] technology over the next 5 to 10 years?

[00:27:35] I think it’s the same thing that

[00:27:36] separated them over the last 5 to 10

[00:27:38] years. An inherent curiosity, an

[00:27:40] inherent sense of urgency and a passion

[00:27:43] for it. And those things you don’t

[00:27:44] teach. You know, you you see the people

[00:27:45] that that come with it and when you see

[00:27:47] it, you need to nurture it. But you

[00:27:49] know, those things if if you look at the

[00:27:50] folks that have been really successful,

[00:27:53] they have those innate qualities. The

[00:27:55] other things you can teach, you can

[00:27:56] learn. But if someone doesn’t have that

[00:27:58] sense of urgency, that that innate

[00:28:00] curiosity, they’re not going to move

[00:28:01] forward. and and right now being curious

[00:28:04] is really what AI is all about and

[00:28:06] whatever the next iteration of AI is

[00:28:09] which I I don’t even know what to call

[00:28:10] it at this point but you know we’re just

[00:28:12] going to see all of the blending and

[00:28:14] becoming native in there and it’s that

[00:28:16] urgency creativity you know and just

[00:28:18] curiosity it’s just so vital. if you had

[00:28:21] to leave our listeners with one hardle

[00:28:23] learned lesson about balancing

[00:28:26] innovation and the foundations

[00:28:28] underneath it or perhaps from your life

[00:28:31] what would that be and why

[00:28:33] >> so I’ll actually make two points the

[00:28:34] first advice I I give everybody is

[00:28:37] follow your passion the technology

[00:28:38] career is a long career there’s a lot of

[00:28:40] highs and lows there’s a lot of crisises

[00:28:42] everybody who’s who’s in technology will

[00:28:44] experience a crisis whether it’s a cyber

[00:28:46] or an operational crisis and if you

[00:28:48] don’t love what you’re doing you’re not

[00:28:50] going to last, you’re not going to enjoy

[00:28:51] it. Um, you know, the the the other

[00:28:54] point is always technology is moving

[00:28:57] faster than ever, but always make sure

[00:28:59] that you slow down just enough to plan

[00:29:02] to to think it through. It is right now

[00:29:04] because we move so fast, moving in the

[00:29:06] wrong direction too quickly without some

[00:29:09] careful thought. Whereas before you you

[00:29:11] would only move 10 m down the road, this

[00:29:14] technology moves so fast that if you

[00:29:15] make a turn, you’re moving 200 m. And so

[00:29:17] now that course correction is a lot

[00:29:19] tougher to it. So it is okay to take a

[00:29:22] breath to slow down and say let me think

[00:29:24] about this. Let me ask my my my peers to

[00:29:27] it. It doesn’t mean that you don’t know

[00:29:28] what you’re doing. It’s not a sign of

[00:29:29] weakness. It’s actually a sign of

[00:29:31] maturity with it. Yes, you can go and

[00:29:33] vibe code this amazing product faster

[00:29:36] than ever before, but before you put it

[00:29:37] into production, make sure you’re

[00:29:39] checking it. Make sure you’re doing the

[00:29:40] right things. That will pay off 10 times

[00:29:44] the effort. an extra minute on on

[00:29:45] checking or you know as carpenters like

[00:29:48] to say you know measure twice before you

[00:29:51] cut that still is an adage I don’t care

[00:29:53] how much things are moving now obviously

[00:29:55] the art is do I measure 10 times or do I

[00:29:57] measure two and I think that’s the art

[00:29:59] that we have with within technology is

[00:30:01] is figuring out that right balance and

[00:30:03] the balance is different for every

[00:30:05] industry for every tool and for every

[00:30:07] leader

[00:30:07] >> but what among these two advices was the

[00:30:10] hardest learning curve for yourself

[00:30:12] >> the latter I’m hyperactive I like to

[00:30:15] move quickly and sometimes I have to

[00:30:17] stop myself or be honest with you I have

[00:30:20] some of my staff will say shouldn’t we

[00:30:22] do this an extra test or shouldn’t we do

[00:30:24] this and I’ll go yes you’re right and I

[00:30:26] still suffer from it today because we

[00:30:27] want to innovate so fast sometimes we

[00:30:29] want to crawl run instead of crawl walk

[00:30:31] run and that is a lesson that I’ve

[00:30:33] learned and a lesson I have to remind

[00:30:35] myself all the time

[00:30:37] >> no thank you so much for your grounded

[00:30:39] and insightful conversation