Syntax - From AI First Thinking To Data First Reality

[00:00:03] Welcome back to the Tech Talks Daily Podcast. At the moment, I'm recording 60 podcast interviews a month across the eight podcasts I now host on the Tech Talks network. And somebody asked me recently, hey Neil, who does all your editing? How do you do the marketing, the production, everything? Let me tell you. I'm going to give you a little secret. It's all me. There is no special team. Everything you see and hear is me. So a big thank you for choosing this podcast to listen to.

[00:00:33] And one of the things I want to talk about today is what is really holding AI back. Is it those algorithms that we choose? Or is it the data that we trust them with? And today I'm joined by an incredibly cool guy. I live in the West Midlands in Wolverhampton, which is very close to Birmingham. And my guest today, Kevin Datilico, he's from a company called Syntax. He's talking to me today from Long Island.

[00:00:58] But he did come over to Birmingham for the Aussie Osborne farewell tour last year. So I was able to have a good chat with him about music, the area and everything in between. But today, the focus of our conversation is going to be about unpacking why so many AI initiatives hit a wall just as they move from promise into production.

[00:01:20] And he works with organizations running some of the most complex mission critical systems in the world across SAP, Oracle, cloud platforms and everything in between. But most importantly, I want to dig into what he's seeing right now. And one of the things he's going to talk about is a pattern that many leaders will recognize. And that is, yeah, AI pilots, they look impressive. Proof of concept? Yeah, they get all the applause.

[00:01:50] But the reality arrives the moment those models touch live operational data. So yeah, we will cover that age old phrase, no data, no AI. Why it keeps showing up at conferences and why it is so much more than a slogan. Because Kevin will explain where the biggest data gaps emerge when structured ERP data collides with the messy unstructured world of operational data.

[00:02:15] And why pushing ahead with weak foundations creates operational compliance and reputational risk problems. Probably best to avoid that. So if you are a business leader feeling the pressure to justify AI spend while quietly realizing that your data house isn't in order yet. And from the conversations I've been having, I think there are quite a few of you out there. So you are not alone. And today's episode will help you rethink where real progress can actually start.

[00:02:43] Yep, we are solutions, not problems. Lots to learn today. But enough from me. Let me bring Kevin onto the podcast now. So a massive warm welcome to the show, Kevin. Can you tell everyone listening a little about who you are and what you do? Absolutely. Thank you. My name is Kevin Datilico. I'm the regional CEO for the Americas at Syntax.

[00:03:06] And essentially, I work with organizations across multiple industries, really helping them modernize and run their mission critical systems. And this falls into the SAP, Oracle, and cloud platforms. And my focus is really trying to help customers, I'll say, simplify complexity, unlock value from that data that's within these systems, and build technology foundations that support long-term innovation.

[00:03:34] And that obviously now includes AI. Yeah, that AI word. We've lasted only a few minutes before mentioning it. But it is a tech podcast. We've got to go there. And there's so much I want to talk with you about today. I am in the US a lot for tech conferences. And if I was to take one key theme out of all the tech conferences over the last 12 months, one phrase that I heard repeated many times was not agents or agentic AI, was no data, no AI.

[00:04:02] And one of the reasons I bring that up is because I was reading that you've said that many organizations will soon pause or be forced to recalibrate AI initiatives once they realize the limits of their data. So what is it that usually triggers that moment of realization for leaders? Yeah, what I would say is, and I'll say unfortunately, it usually happens the first time the model hits production.

[00:04:28] And what happens is that leaders see an early proof of concept win, but the reality is, is that once AI has the ability to interact with real-time operating data, and that's data that's inconsistent, duplicated, or locked in other legacy systems, they realize the results are unpredictable.

[00:04:52] And they start to see hallucinations, conflict in outputs, or KPIs that really don't reconcile with what the business needs to see. And hopefully, and sometimes results differ, but that's when the organization should understand that AI isn't the issue, it's the data foundation. Some people blame AI, but it's not going to fix all those things. Ultimately, it's the data foundation.

[00:05:23] And I think traditionally, ERP data has been seen as somewhat as reliable, but unstructured data across the enterprise often tells a very, very different story. So where do you see the biggest gaps when companies try to bring those two worlds together? Any big surprises that you see there typically? I don't know if it's surprises, but I do think it comes down to really three key areas. And I think the first one is context. And you know what?

[00:05:52] ERP data is very structured, process-driven, but unstructured data is highly nuanced. It's within emails. It's within engineering files. It's within images. It's in IoT logs. And without context, the systems don't agree. The second one I would say is granularity. And ERP data captures transactions.

[00:06:21] Operational data captures behavior. And bringing those two together requires alignment at a very, very detailed level. And I think the last one, and to me it's probably the most important, is that the ERP has clear governance. Unstructured data within many of the organizations doesn't have that governance.

[00:06:46] And you can't unify data if no one owns its quality. And as businesses begin or continue to combine ERP and operational data with the best of intentions to support knowledge workers, etc., what is it that typically breaks first? Is it the technology, the processes, or even the assumptions teams make about their data quality?

[00:07:14] What are the warning signs that usually manifest first? So I think it's the assumptions break first. And it's usually long before the technology does. And most teams assume that since data exists, it has to be accurate, it's aligned, and ready for AI consumption.

[00:07:36] But once they start stitching the systems together, they discover it's conflicting hierarchies, outdated master data, disconnected processes. And technology, I think, today can bridge almost anything, not everything, but almost anything, but only if the underlying processes and assumptions are grounded in reality.

[00:08:04] And I think after decades of working with data silos, poor quality data, seeing the results of garbage in, garbage out, there is thankfully a growing belief that AI and analytics outcomes are dictated by the strength of the data underneath them. So at least we're beginning to understand this. But for business leaders listening, what should they be doing to rethink investment priorities when budgets are already committed to AI tools? And maybe they were blindsided by the data issue.

[00:08:35] Yeah, yeah. And that's the tough part because AI is the key thing right now. And I think those technologies have outpaced what's happened from a data side. So what I always try to preach is that leaders need to shift from an AI first to data first. And that doesn't mean to really obviously abandon the AI investments, not at all.

[00:09:01] It just means ensuring that those investments produce business value. And I think the most successful organizations rebalance the spending towards one, data integration and consolidation, master data management and quality automation, real-time ingesting pipelines, and meta lineage and governance frameworks.

[00:09:27] With all of that, those AI investments become a reality and produce the business value that they're expecting. Oh, absolutely love that. And before you came on the podcast, I was also reading that you mentioned a renewed focus on consolidation, cleansing and transformation all on a massive scale. But again, for that business leader listening, what does that look like in practice for their large enterprises when they're faced with years of fragmented systems?

[00:09:57] Exactly. That's a tough road. So it has to be a multi-phase effort. Honestly, you need to be able to rationalize the systems and retire the technical debt. And that technical debt could be a heavy, heavy burden that somebody's carrying along, but we have to work towards retiring that debt. And that involves streamlining redundant platforms. You really need to industrialize data cleansing.

[00:10:27] It's not a one-time cleanup, but it really has to transform to an automated pipeline that maintains quality continuously. And this one almost may seem a little obvious, but it's amazing what happens within these large organizations. And it's really standardized definitions. You know, we all need to agree on what a customer product and location actually means globally.

[00:10:57] That sounds simple, but it is so, so important, especially as we're ingesting all this data. And then I think the last one is really to build centralized data products that allow you to curate and govern data sets that feed AI and analytics and operational workloads. And to that business leader listening now, yeah, it will feel like a heavy burden at the moment, may feel like an incredibly overwhelming task.

[00:11:25] But I think it's also important to highlight there is light at the end of the tunnel. And as data becomes more centralized, governance and security often become friction points as well along the way. And maybe that's something else we might need to address or think about addressing. So how are organizations redefining governance and role-based access without slowing decision making? Because it feels like a real period of transformation here, but with the right decisions on that journey, so many big opportunities here too, right?

[00:11:55] Absolutely. And I do love to talk about the role of the chief data officer. And I think this one fits in this one perfectly because the chief data officer traditionally was always put in place to restrict and to govern. And it was done for compliance.

[00:12:16] And now this role, I see it as a massive transformational role in organizations today because now they're innovating and enabling with guardrails. And that becomes really a key piece. And what I think here is what has to happen is that there needs to be a federated governance model in which there's central standards but local flexibility.

[00:12:42] It's not just one centralized governance model. And it has to be an attribute-based access control instead of static roles. And there needs to be automation for classification of masking and lineage so governance is embedded and it's not manual because today it winds up being a heavy, heavy manual lift.

[00:13:06] And then I think most importantly, you have to get to the point of having a data marketplace where users can request or subscribe to data products with clear guardrails. So as you said, to me, this is the biggest transformational area to allow us to be able to really realize the benefit of AI and analytics.

[00:13:32] And I'm curious, from the work that you're doing and the conversations you're having with customers, what are the risks of pushing ahead with AI on top of weak data foundations, especially in regulated or complex industries? Because I would imagine that you've got a lot of war stories under your belt and seen a lot of mistakes made. But what are the big things you've picked up here? What happens if you ignore what we're talking about? Yeah, so I think it really falls into three categories.

[00:14:02] One, I think there's an operational risk. And AI models that generate unreliable insights honestly drive poor decisions. The second one is a compliance risk. Bad data mixed with automated decisioning can trigger audit issues or even regulatory violations.

[00:14:24] And one that's probably potentially the most critical one and probably one that's difficult to recover at times is really a reputational risk. In industries like manufacturing or healthcare, inaccurate AI recommendations can impact safety, quality, customer trust, or even life decisions. So that reputational one is really, really a difficult one to recover from.

[00:14:53] So extremely important. Quick thank you to the sponsor supporting all of the shows on the Tech Talks network. And this month I've partnered with Alcor. And if expanding engineering operations beyond your home market can be overwhelming, you're not alone. Because if you've ever wrestled with local laws, slow response times, and partners who treat each country as separate rather than part of a wider strategy, you might want to check out Alcor.

[00:15:22] They approach expansion completely different. They specialize in building tech teams across Eastern Europe and Latin America. And they combine employer of record services with recruiting. So you get one singular coordinated process. They help you choose the right jurisdiction based on your needs, run proper evaluation of candidates, and onboard teams quickly. And their model is also refreshingly transparent. Most of your contribution will go straight to your engineers.

[00:15:51] And their fee shrinks as your team grows. And there is no cost to exit if you move the team in-house at a later date. And I think that kind of clarity is why so many high-growth companies in Silicon Valley are working with them right now. So you can find out more details at alcor.com slash podcast or simply use the link in the show notes. And as I said a few moments ago, there are so many opportunities here.

[00:16:20] And I'm a solutions, not problems kind of guy. So looking forward in the weeks, months, and maybe even years ahead, how do you see organizations that listen to the advice that is being offered here today and strengthen their data core, pull ahead of their peers when it comes to analytics, automation, and future technology? What's the big payoff? What is the exciting opportunity ahead there for companies that do this?

[00:16:47] Well, so first, I think the leaders that treat data as an operating asset and not as an IT byproduct will pull ahead. And they'll be able to honestly deploy AI faster with more predictable results. And I think that's the important piece. They can automate processes end-to-end instead of just isolated products in which they know they have that clean data foundation.

[00:17:17] They could go end-to-end. They can now adopt emerging technologies like autonomous systems, edge AI, digital twins, without re-engineering everything underneath. And one of the most important ones, too, is that they can begin to trust their data because then that allows accelerated decisions.

[00:17:41] And it reduces the oversight cycles that you constantly have to be monitoring over to make sure that it's given us the right responses that we expected. So I think strengthening the data core becomes a force multiplier. It doesn't just support AI. It really reshapes how the business operates. And I'm curious, from everything that you're doing, all those conversations, again, with all your customers scattered all around the world,

[00:18:06] is this a big focus for you at the moment at Syntax or any other big trends that you're working on or excite you at the moment? What's taking up all your time? So to me, I do really have AI as the – or I should say data is the new currency. And AI, analytics, and automation are really just force multipliers for that data structure.

[00:18:34] So right now, most of my time is spent with taking customers through the rest of their data journey and how they can actually truly realize the business outcomes that their leaders are asking them to deliver. So to me, it's very exciting. It's having all these technologies mature and the capabilities that are there.

[00:19:02] In my career, it's probably not all these variables have lined up just in the way that they are right now. So I think it's really, really a tremendous, exciting time. And if you put all those conversations into one big melting pot there, are there any trends around the kind of things that people are struggling with and asking for your help with? And the reason I wanted to ask this is I think there's a lot of people that have been listening around the world and they almost begin to feel that there's only them struggling with this particular problem.

[00:19:32] But for my conversations I've had on show floor at tech conferences, et cetera, everyone's talking about the same thing and the same problems. Is that what you're seeing as well? And that one always makes me smile a bit. Look, just as human beings, we all think we are so different and we are different in so many different ways. But you know what? Once we actually start talking, we realize, you know what? We're really not that different.

[00:19:59] So we have a number of customer forums in which we bring customers together from all different industries and utilizing all different technologies, whether it's SAP, whether it's Oracle or whatever it may be. And they all think they're different based upon the technologies that they're using. But in reality, they're all trying to solve the same problems. They're using different technologies.

[00:20:22] And it's really the power of whether it's engaging with a partner or engaging with a peer and a customer and exchanging ideas. And for us, when we bring these customers together and they're really exchanging those ideas, I can't tell you the number of times that I get feedback from the customer saying, Kevin, thank you so much. You've put me into a safe environment in which I can openly talk about my problems without any fear or repercussion of something's going to happen.

[00:20:49] And we're able to exchange and see what worked for me, what didn't work, what did I try? And it's a true learning thing. And once we allow ourselves to be a little bit more vulnerable and share those things, then I think everything starts moving forward faster. And I think that is a powerful moment to end on. And we have been very forward-looking today, talking about you, your career, everything that you're working on, not just today, but looking towards the future.

[00:21:17] But before I let you go, I'm going to ask you to look back now at your career. And the reason I say this is I think very often there's somebody early in our career that sees something in us and invests a little time in us and play a big part in our career. And they're blissfully unaware of it. So is there a particular person that you're grateful towards that we could give a little shout out to today?

[00:21:39] Yeah, it's funny that you mention that because I do a thing with a group of my friends that I've been friends with since elementary school, like over 50 years. And one of my friends asked me a very similar question.

[00:21:53] And the answer I gave him really surprised him because it was one of my best friend's fathers who, during my high school time, allowed me to take part in a – at the time, he worked for a large bank in New York. And they allowed family members to intern. And it was really only open for family members. And I was just a friend of his son's. So we always got in trouble together.

[00:22:22] And him and his wife really took me under and said, hey, look, we want to give you this opportunity. You're as close as family. But remember, you're representing us within this company. And from that point forward, I was immersed into technology from a banking perspective. And it carried forward into college. And eventually, I worked there full time. But I can actually point to every single role and job I've had starting from that point right there.

[00:22:51] And my friend was in awe because he expected me to say some other big, great business leader that influenced and helped me. And he said, have you told him recently of the impact that he had on you? And I said, you know what? I haven't seen him in a while. And he just turned 85. And I know his hearing's gone or whatever. I said, you know what? I really need to take the time to send him a letter and really express that and really thank him.

[00:23:19] And you just never know where somebody is in their own personal and professional journey. And just that message alone. I know his family shared it with me. He was very emotional. And it just really hit home. And it's just the simple act of acknowledging somebody and really thanking them and just how much they really meant and impacted you and your career is such an impactful thing. So we should always take that time and look back.

[00:23:46] So I really appreciate the question because it was meaningful to me. And it's meaningful to the individuals. So before we only have so much time on this floating rock. So take the time to appreciate and thank the people around you. And it has such an incredible impact. So, yeah. Oh, man. In 4,000 interviews, you're the first guy ever to bring a tear to me. That is just a real beautiful story.

[00:24:14] And I think it's so important to recognize these people that have such an impact on our lives. And for the most part, they're blissfully unaware just thinking you've gone to better things, you know. Exactly. Changing the subject quickly. You don't want to cry in podcast here. So before we started recording today, you were telling me that you visited the UK earlier last year to go to the Aussie Osbourne farewell concert, should I say.

[00:24:42] And I've got a Spotify playlist. I don't know if we've got any Osbourne or Black Sabbath on there. And as a rock and roll rebel, what song would you like me to add to that playlist? I'm going to let you choose one. What would you like me to add and why? No, I would put Crazy Train on there. So many people know that song, but if you really look at it, Aussie gets this reputation as being this wild, crazy guy. But if you really look at the lyrics, you could classify them as an older hippie type.

[00:25:11] So even if you listen to the words of Crazy Train, it goes millions of people living as foes. And we need to learn how to love and forget out of hate. And it's like, even through all the crazy guitars and everything else, the message that's in there is just so powerful and just so relevant through everything. Especially, you know, you turn on the TV and see all the things going on today.

[00:25:35] And it's just like, seriously, we need to focus more and trying to focus on more love and less hate. And we'd be much better off. So, yeah, that's a beautiful moment to finish on. And before I let you go, finally, we've referenced reports today, a lot of stats, a lot of big information. For anyone listening wanting to carry on this conversation, connect with you or your team, reference any material, keep up to speed with things. Where would you like to point everyone listening?

[00:26:05] Absolutely. Obviously, you go to www.syntax.com to understand what we're doing there. Always feel free to reach out to me via LinkedIn, send me a message and be happy to continue the conversation. Because like I said, it's a really exciting time and it's a topic I'm really passionate about. So I will add links to absolutely everything. I urge everyone listening to check that out. But more than anything, just thank you for joining me today. Absolutely. I really enjoyed it. Thanks again. Next time I head out to Birmingham, I'll come find you.

[00:26:35] I think one of the big takeaways from this chat today is that organizations pulling ahead are the ones that are treating data as an operating asset, not just another IT byproduct. And when data foundations are strong, then AI can become predictable, scalable and trustworthy. On the flip side of that coin, when they're weak, everything slows down. Decisions need constant oversight and confidence slowly erodes.

[00:27:04] But this episode today didn't just land on technology and transformation. It landed on something far more human. And near the end of our conversation today, that moment where Kevin shared a deeply personal story about someone who quietly shaped his career. A figure who opened the door at exactly the right moment was simply beautiful.

[00:27:26] I think what made it even more powerful was what happened next when he took the time to sit down and write a letter of gratitude many, many years later, simply to say thank you. And I think that letter meant more than even probably Kevin understood. But the person who received it, I think was just next level. And honestly, I did have a tear in my eyes. He told that story. And it certainly made me think who I should be thanking.

[00:27:53] It was helped me in my life and in my career. And maybe if it struck a chord with you, maybe you could sit down and write a letter to somebody who you're eternally grateful for. For a few moments that they spent with you that were there for you, pointed you in the right direction. For the most part, people are blissfully unaware of how grateful that you are. But before I set off again and get all teary on you, it is a tech podcast.

[00:28:19] And this discussion was about AI, data and modernization. And that moment, I think, was the reminder that progress is still built on people, on trust, on taking time to acknowledge those who've helped us, even when they didn't have to. So you will find links in the show notes to syntax and the topics we discussed today. And I'll leave you with this question.

[00:28:43] As you push forward with data, AI and transformation inside your organization, who is the person that you should be thanking for helping you get here? Let me know your thoughts. Techtalksnetwork.com Send me an audio message. If you've got a story you want to share, go to techtalksnetwork.com Go to contact. Hit record. Send me a little message. I'd love to hear it. Be nice to hear your voice rather than you just hearing mine day after day. But that is it for today.

[00:29:13] So thank you to Kevin for an interview that I won't be forgetting in a long time. And my thank you, well, that's to each and every one of you. Every single day you turn in and allow me to be the soundtrack to a few minutes of your day. So thank you. And on that note, I will bid you farewell and I will speak with you all again tomorrow. Bye for now.