Why Data Governance Should Come Before AI According to ProArch

[00:00:00] The leading issue of agentic AI in businesses right now is ensuring agents act with compliance guidelines. And Denodo applies guardrails across your entire data estate. By aligning your company's data infrastructure under one system, these guardrails perform consistently across your platform. So start scaling your business and start with Denodo. Simply visit denodo.com to learn more.

[00:00:30] Welcome back to the Tech Talks Daily podcast. I'm recording this at 4pm on a Friday here in the UK. Low grey skies overhead. But I'm talking to someone in upstate New York who knows a thing or two about snow, lake effect weather and long winters. But today we're not here to talk about the climate outside. We're here to talk about the climate inside the enterprise when AI shows up.

[00:00:59] And let's face it, it is everywhere right now. And here's the thing. Everyone says they're doing AI. But operationalizing AI inside a large organization without creating blind spots, that's a very different challenge altogether. So today I'm joined by Jim Spignardo, Director of Cloud Strategy and AI Enablement at ProArch.

[00:01:23] He spent 25 years across networking, cloud, cybersecurity and now AI strategy. But his focus is refreshingly practical. It's about turning AI hype into measurable business outcomes. It's about data governance, technical debt and the uncomfortable truth that you cannot layer AI on top of broken processes and then simply expect the magic to happen. It doesn't work like that.

[00:01:50] So today we will get into what AI enablement really means beyond switching on Microsoft 365 Copilot and how to embed AI into workflows instead of treating it like just a shiny add-on. And most importantly, why augmentation, not replacement, is the mindset that wins in the long term.

[00:02:13] So if you're leading IT, shaping strategy or just wondering how to show real return on AI investment, this conversation will be packed with insights that you can actually use. But enough for me. Let me get Jim onto the podcast now. So thank you for joining me on the podcast today. Can you tell everyone listening a little about who you are and what you do? Sure. My name is Jim Spignardo.

[00:02:39] I am the director of cloud strategy and AI enablement for ProArc. ProArc is a global organization. We do a lot of different things, mostly focus on data, AI and security. We have managed support customers throughout the United States and we also operate our own 24-7 security operations center as well. Well, thank you for sitting down with me today. There's so much talk around AI. There has been for the last three, four years now.

[00:03:08] But although people talk about how transformative it is, you and I have been around the block a time or two and seen many different trends. I mean, you've spent more than 25 years across networking, cloud, cybersecurity, now AI enablement. So when enterprises talk about operationalizing AI, where do you see them creating new blind spots without even realizing it? Because we've seen in the past what happens when Silicon Valley moved fast and broke things, but we don't want to repeat that.

[00:03:37] But do you see any blind spots coming out? One that scares me a little is the agent thing. You know, we're talking about hundreds of agents out there. That makes me a little nervous. But what about yourself? I think agents are definitely something to think about and consider risk. The big thing, I think, from my perspective, though, that a lot of organizations are not factoring into their journey is what they're doing around their data governance.

[00:04:02] And I'm actually going to be putting out an e-book in the next few weeks on this topic of if organizations don't finally deal with all the legacy debt, technical debt that they have in their organizations, they're going to really find it very difficult to operationalize AI in the ways that they hope to.

[00:04:20] They can't just rely on old, broken processes that have been moved along through sheer will and effort over many years and expect them to be converted into something glorious and automated without dealing with who's responsible for that data, who owns the data, making sure that data is cleaned and in a pristine state that allows you the best possible chances of success when you're applying artificial intelligence to it.

[00:04:49] You know, it's no different than, you know, asking someone to come in from that's never seen your organization again and trying to make and understand what it's all about. You know, if you really think about how AI functions, it's no different than how humans function. If the environment is complete chaos, well, expect that the answers are going to be a little bit squirrely and a little strange until you can get things in your house in order and your data in shape.

[00:05:13] And one of the reasons I invited you on the podcast to join me today is when I was doing a little research one day, I was reading how you often use the phrase AI enablement. So does that actually mean for IT leaders, strategy teams, broader workforce and people listening beyond simply just deploying a model or turning on a co-pilot feature?

[00:05:33] Yeah. Yeah. I mean, because well, well beyond deployment where enablement is, is really embedding it into the workflow, the work processes of your organization and really taking a different thought process about how work gets completed. And, you know, I often say to a lot of customers too, because they come to us and go, what tool should I use? It has nothing to do with the tools, right?

[00:05:57] It's really about understanding how you want to leverage these, this technology to transform the business. And enablement really is how do we begin to ingrain these tools and these platforms into how work gets done and how business gets accomplished. And that is really about changing minds and attitudes and also work habits, right? That's goes well beyond just saying, okay, here's a new tool to do something cool.

[00:06:26] It's really about a yesterday we used to do all of these things that took us a lot of time and effort, distracted us from real work. Today, we are able to leverage AI and automation to be able to take that off our plate so we can focus on really generating value for organizations. And at ProArch, just to bring you guys into that, you bridge cloud infrastructure, data architecture, and human readiness.

[00:06:50] So how important is foundational cloud and data maturity before organizations attempt to scale AI initiatives? And one of the reasons I ask that is one of the phrases I keep hearing at tech conferences is, hey, no data, no AI. It's kind of table stakes, isn't it? Yeah, absolutely. And these are conversations, honestly, that we've been having with customers well before the advent of ChatGPT or Generative AI.

[00:07:14] But it was so much harder to get companies to take it seriously, to understand that managing your data estate really gives you the opportunity to be able to extract very good insights from that data. And so when we look at, you know, we can deploy things like Copilot from an end user productivity state, and it's not that big of a deal.

[00:07:38] But when we start talking about deploying cloud platforms and developing our own solutions that are going to connect to back-end line of business systems and integrate with them, our digital engineering team really pumps the brakes and says, we're going to do a data assessment first. And we're going to make sure that if the data is not in a pristine state, we're either going to move it into a data platform where we can clean it and get it ready for these tools,

[00:08:03] or you're going to have to spend some time really going through and ensuring that there isn't legacy data in there, there isn't erroneous data, old outdated data. The data has the appropriate tagging and structure to it so that it can be properly reasoned over. So yeah, it's very, very critical.

[00:08:23] And as I mentioned, those that can kind of begin to put their arms around this and really be honest with themselves as organizations about what they've been ignoring for all this time are really going to be the ones that can succeed. And it's interesting to see now that we've had, now that AI is here, how many of those customers that we talked with a few years ago are coming back and go, hey, let's talk about data governance. Oh, now you want to talk about data governance. I see.

[00:08:52] Absolutely love that. And we will have many people listening that have had mixed experiences with 365 Copilot. So as ProArch's internal champion for Microsoft 365 Copilot, you went on to build an AI adoption playbook, which is incredibly cool. So what are the common mistakes that you see organizations making when trying to translate AI excitement into real workflows? Tell me more about that. It's work, right?

[00:09:19] So we talked about, you know, you can't just implement these tools and expect people to just gravitate toward them and understand how to use them. Some people will. And luckily, we're a technical organization. We have a lot of brilliant people. But what we really chose to do is start with identifying use cases and looking at the personas and roles of their organization that would benefit. We took a slow approach. We onboarded a different role each month over the course of a year or so.

[00:09:49] And by doing that, we introduced those teams to the technology. We asked them, you know, where are the pain points today? I always like to talk about the three D's of that AI can address. The dull, the distracting, and the draining, right? So if you think about it from that perspective, what type of tasks do you do today that fall into those categories that we can use these tools with?

[00:10:14] What we came to understand, even in our own journey, was you can't just assume you've created a use case now and you've given them the playbook of how to use it. You have to make it mandatory. You have to explain this is how you will now perform these tasks. Because people are very apt to just go back to their old ways and say, well, I know you told me to do this. I really like taking notes by hand.

[00:10:38] Again, I'm not going to stop someone from doing that, but if the direction is all meetings will now be transcribed and those notes will be saved to our system of record, right? There has to be some level of ownership within that team to say, yeah, and we're going to be doing this 100% by this day.

[00:10:57] So it's making sure that you have metrics to track, you have KPIs that you've built to also demonstrate how you're moving the needle as it relates to the adoption of the technology and the value that is returning to the organization. And we spend a lot of time also just with continuous training. We have a center of excellence, which is a knowledge hub where people can go and find information about these tools.

[00:11:22] I just got done doing today our monthly user group, internal user group, where we pick a topic every month and we do a workshop and we get to get hands on with the technology. We have a weekly newsletter and a weekly tip of the week that we put out. And we also have champions that are available to the teams to, if they have an idea, they can come to them and work through it to see how to actually bring that to life.

[00:11:48] And we will have many people listening and worried about AI replacing jobs. So from your client work, all the conversations you're having, the people you're working with, how are organizations successfully using AI to augment human decision making rather than sideline it? Because I think there's enough of that bad stuff out there. We've all seen it in our news feeds, but to restore a better balance in the universe, it'd be great to hear some positive stories.

[00:12:13] I think the first thing is you need to have it from the executive level, a statement of what the intent is with these tools. So we were very intentional when it came to making sure that our employees understood we're using this, as you said, to augment their capabilities, not to replace them. And so if I go back to the concept of the three Ds, we want you to get rid of those things out of your workday so we can really use you to your fullest potential.

[00:12:42] And from a creative standpoint, from an ability to get in front of our customers more, to talk to them face-to-face, we're aware of having you fill out timesheets or create financial reports every month. That's not the best use of your time. The technology exists today to be able to kind of remove that burden from you so you can do more creative thinking and more solutioning of how we can grow the business and continue to prosper.

[00:13:08] You're someone that advises CIOs on aligning IT with business outcomes, which is so important. Over the last couple of years, we've seen the big ROI topic resurface. So how should leaders measure ROI on their AI investments in a way that resonates with finance and the board? It is such a massive topic right now. You must get people asking you this a lot. But how do you see this playing out?

[00:13:33] Yeah, and it's a topic that we talk about in our executive briefings that we do for organizations that are just getting their, stipping their toe in the water. The important thing is, initially, you will be tracking just general usage. How many people are using it and how often are they using it and what types of tasks?

[00:13:50] But to really get to the point where it's going to be something that is transformational for the business, you have to start looking at, again, within the use cases themselves that you've developed, based on those metrics that you define that you want to track. How do we translate that into a dollars and cents number?

[00:14:10] So, for instance, if I'm talking about a sales cycle within an organization and we have an agent that helps build proposals for us based on pre-sales meetings. We can take the notes from a pre-sales meeting and generate a first draft within an hour after the meeting is completed. We want to know how does that help the business? Are we able to put more proposals in front of clients? And if so, does that mean we're winning more deals as it's actually moving the needle as it relates to revenue?

[00:14:40] Not everything is going to come back to revenue, but those are the ones that resonate the most, right, especially at the C-level. So, if you're an organization that wants to really kind of be able to continue to drive that adoption, it's important you find things that do translate back to dollars, right, that actually show those folks in the C-suite that this is paying for itself. So, it's a no-brainer. Let's just pay for the technology because we've demonstrated it's helping us here to be more productive.

[00:15:09] It's helping us here to win more sales. It's helping us here to hold headcount, not necessarily decrease headcount, but hold headcount. And those things are going to win the day every time. You know, we went from a pilot with about 10 licensed users initially to 18 months later having about 150 of our 500 employees licensed for Copilot. The only way we could do that is by demonstrating to that executive team that we were getting return on value.

[00:15:37] We started around 2x, and now we're somewhere at 9.5x, and it continues to improve. Awesome. So good to hear stuff like that. And also, one of the things I always try and do with my guests is give them a virtual soapbox of sorts to later-esque any frustrations around myths and misconceptions that they may be seeing on LinkedIn, Reddit, or wherever they get their daily tech news.

[00:16:01] I'm sure as someone working right in the heart of this space that you must see a lot of news or stories that you'd completely disagree with or you just know are not factually correct. Are there any myths or misconceptions about your work that you want to lay to rest today? Does anything stand to mind? Yeah. There is. And, you know, I have three children who are Generation Z, right? And so they are naturally skeptical based on just growing up in the world that they did. You know, they were digital natives.

[00:16:31] But I got to tell you, they have a tremendous amount of skepticism as it relates to what AI is going to do. And one of the things I think we're just unable to put into context is what these tools can do and what it's going to mean for society long term. And I think everybody immediately goes to the doom and gloom scenarios. There's never going to be live music again, right? There's going to be art classes and paintings. And to the extent, will that change? Absolutely.

[00:16:59] I mean, if you look at the movie industry, how many movies nowadays are not, don't have some level of CGI in it? Yeah. And there was a little bit of a resistance initially. I'm a James Bond fan. And so when I saw CGI, because it was bad CGI too, and some of the first movies there, I'm like, is this where we're going? This is terrible. But it got better. It got better and it became more natural. And it didn't replace actors.

[00:17:25] It kind of augmented the abilities to tell stories in a little bit more vivid and creative ways. I think AI is going to do the same thing. And we need to kind of put it in context and realize all of these things have a way of working themselves out over time and becoming more accepted as part of how we just get things done.

[00:18:15] Love that. And we're going to do this journey. They're going to go very fast and going to seem like, wow, we're right out on the edge of the future. And then, like with every technology, there are certain industries that are going to take a very long time and probably will not necessarily be hurt by not adopting this stuff right away. Right.

[00:18:38] If I'm in a lawn mowing business or I have a highway construction business and if you're not adopting AI tomorrow, I don't think you're going to be put out of business by the guy who is. Yeah. But there are definitely industries that are going to differentiate themselves by the ability to leverage these tools to be more efficient, more productive and outpace their competitors.

[00:18:59] If you look at what Walmart's doing, they have this super agent with their operational environment where it's an agent that basically runs all their other agents. If I'm someone like Target or some other big retail chain, that concerns me, right? How much quicker can they get things done? How much faster and cheaper can they do that stuff than I can?

[00:19:19] And so those areas I think are where we're going to see the most disruption is at that kind of high, the top in enterprise level where they can spend a lot of money and potentially even burn a lot of money initially to try and get those things right to compete on a global scale. At the very beginning of our conversation, you mentioned you were working on an e-book as well. Tell me more about that. When can we expect that out? What's it about? What are you doing? Assuming I get my act together.

[00:19:46] That should be coming out probably beginning of March and it's called the AI Turning Point. And it really deals with business and business process and the old ways things were done, kind of what we mentioned at the beginning of this conversation and how that's not going to be sufficient anymore if we're going to put these tools into our businesses. We have to build more ownership.

[00:20:10] We have to be more honest with ourselves and organizations about what works and what doesn't work and what we've just made work through just sheer force and effort. And because ultimately if we try to pivot to this technology without doing that, the results, I'm not going to say could be disastrous, but they're just going to make things more difficult for people, not easier. And if you get to that situation, people start to blame the technology and they'll resist it.

[00:20:39] And now you're in a situation where you got a lot of repair to do on a psychological level. A good example I can give you is I worked in healthcare for the longest time and we're moving to a new electronic medical record system. And they decide to use dumb terminals or thin clients to support the back end of that. They didn't do their due diligence as far as the infrastructure.

[00:21:03] So the thin clients were seen as the source of all the pain when in reality it was we had a terrible storage infrastructure. They couldn't keep up. We didn't have enough compute. But people came to the point where if they saw one of those things sitting on a desk, they literally would be like, I'm not sitting down. I'm not working on one of those. I'm like, it's not this thing that's the problem. I know you can't see that. And you don't want to get there with AI because it's tough to get past those initial impressions if you don't do it right. I love that.

[00:21:31] Well, I'll be keeping a lookout for that. If you are listening to this in March, I will post date and go back and pop a link in once it is available. But I have taken up far too much of your writing time already today. But anyone listening wanting to learn more about your work, anything we talked about, connect with you or your team, where would you like me to point everyone listening to that? Sure. Probably the best way to get in touch with me is through LinkedIn. I'm very, very active there. I do three articles a week, typically.

[00:21:58] And also repost a lot of the stuff that the company ProArc is doing. There's a link on our website, ProArc.com, where we're going to be doing a webinar on March 4th on building your first declarative agent. So kind of walking people through and showing them a real valid use case that has broad appeal to pretty much any organization. So if the people want to register for that, they can.

[00:22:24] And other than that, like I said, I'm the only Jim Spignardo on LinkedIn, so it should be pretty easy to find me. Awesome. We covered so much today from how to operationalize AI in the enterprise without creating new blind spots, what AI enablement really means for IT strategy and the workforce, and how organizations are actually using AI to augment, not replace human decision making. So I will include links to everything you mentioned there, including that webinar.

[00:22:51] It's something I might even check out myself because I'm quite interested in where to start there. But more than anything, thank you for starting this conversation today. Really appreciate your time. My pleasure. I think one of the big moments that stood out to me from the conversation there was when Jim said organizations have to be honest with themselves about what they've been ignoring for many years. And I think most have a fair amount of technical debt.

[00:23:16] But beyond that, AI has a way of shining a light on legacy data, unclear ownership, and processes that have survived through sheer effort rather than good design. So if you try to automate chaos, guess what? You're going to get faster chaos. But when you take the time to clean up your data, define ownership, rethink workflows, AI can become a force multiplier.

[00:23:43] And I also appreciated his focus on metrics today. Because ProArch moved from a small pilot group to a much broader co-pilot rollout by demonstrating real return on value. That's the language that the board will understand. That's how you shift AI from experiment to strategy.

[00:24:04] So if you'd like to connect with Jim, I'll add links to his LinkedIn profile and the ProArch website in the show notes, along with details above that upcoming webinar that he mentioned. But I'd love to hear from you. Are you embedding AI into your operations or are you still layering tools on top of old systems? And what blind spots are you seeing in your own organisation? Let's keep this conversation going. TechTalksNetwork.com

[00:24:32] You'll find links to all my socials, how you can work with me, leave me a DM or audio message, and the small matter of 4,000 episodes. Yeah, that should keep you quiet over the weekend. But enough from me. I'll return again tomorrow with another guest and I'll hopefully speak with you then. Bye for now.