Subscribe
Share
Share
Embed
Short on time but curious about how intelligent technology and AI are reshaping the enterprise world? Fear not, my friend. The Intelligent Enterprise is your short, sharp pause from the chaos of business as usual.
Every other week, Creative Thinker and Business Strategist, Tom Stoneman, meets with a business leader for a simple, real-world activity outside of tech, to talk about what’s going on inside it. It could be a coffee, a walk, some DIY, who knows? But it’s a chance to take a break, reflect and get to the heart of a single challenge in enterprise transformation with the help of leading experts in digital transformation, change management and workplace innovation.
From AI implementation strategies to debating the next wave of tech leadership trends, this is a show for CIOs, CTOs, and anyone navigating the fast-moving world of enterprise solutions.
The Intelligent Enterprise is brought to you by Digitate, creators of Ignio, the award-winning AIOps platform that keeps tech running smarter, faster and more autonomously. Join us for a coffee, run, or maybe even a dog-walking session, and get the inside scoop on what’s next for the enterprise world, by getting outside of it.
Tom Stoneman:
Hi, I'm Tom Stoneman, and this is The Intelligent Enterprise, where every two weeks we take a break from the chaos of enterprise life and get inside a big idea by getting outside of it. Each episode, we meet an industry expert who helps cut through the noise from all the updates and rollouts while exploring one of their favorite break time activities. It might be over a coffee, a gym session, or even a drive in a classic car, something that gives them some headspace when they're deep in a problem. Today, I'm joined by Matt Harris, co-founder of Darkstar Development. Matt works directly with leadership teams at the point where AI ambition meets operational reality, helping them build the data infrastructure and AI systems that actually deliver on the promise rather than just the pitch.
Matt Harris:
What do we own that nobody else does? How can we be even more successful because we have a market differentiator with this technology instead of just looking at the processes that are already existing today?
Tom Stoneman:
Let's get inside the future of enterprises by stepping outside of them.
Matt, welcome to the show. We're just so happy to have you here. I'm really interested in hearing what you have to say just because of your background. Our focus today really is going to be on why most enterprises are using AI to cut costs when they really could be using it to grow or create something new. There's a quote floating around out there, and I think a lot of people talking about the same thing. I think it was Jensen Huang, CEO of Nvidia, had said that "Companies have been handed the most powerful tool in the history of man, and the first thing they're doing is going and firing people to cut costs." And so I think the take on that is a lot of companies are using AI to do basically the same thing they've always done just at lower costs, just cheaper. And I'd like to get your take on that if you're seeing that out there, because I know your company, Darkstar, really focuses more on development and creating better things using AI.
Matt Harris:
Yeah. One of the things that we really focus on is how do we enhance your overall organization by leveraging AI in ways that make sense? When it comes to the industry, when we look at any sort of a news headlines, you'll notice that there are, there's layoffs happening almost on a weekly basis, and they do get attributed back to artificial intelligence. Now, whether that's the case or not behind boardroom doors, that's to be determined. But some of the most successful organizations that I've seen to be implementing AI has really focused on how they can empower their people to be more productive. For customer-facing clients, how do they support the customers more effectively, not remove the customer service component, yet empower them to be able to support their clients and customers better moving forward?
A huge example of that is something that we were working with an organization where they interacted with a lot of legal entities and a lot of law documentation. This documentation's very vast. They're not looking to get rid of their lawyers. They're not looking to get rid of these legal personnel. They're looking for those legal personnel to be more empowered by giving them a tool that can then bring them back information without them digging through 10,000 PDFs to try and locate that information. So these sort of examples are how we can empower those individuals to be more efficient, be more effective, to make better decisions and not necessarily look at always the aspect of, well, who can I get rid of? How do we empower those that are already there?
Tom Stoneman:
So if we take that into a specific situation, let's put you in a room with a prospect and you've got the leadership team sitting there, maybe even the CEO, but definitely probably the CIO, maybe the CFO, a couple of the senior directors, and you're brought in to talk to them about exactly what you just described. When you sit down with them, what does that conversation ... I mean, what's one of the first things you hear from them that reveals where they are with AI, what their thinking is about AI, and how do you either go along with that or turn it around?
Matt Harris:
So a lot of times, I can gauge that by are we starting with the business goals in mind as far as how does this align to the corporate initiatives? However, on the inverse of that, if we start with the model selection process or we're starting with the only cost savings aspect of it, then I'm very intrigued as to how that aligns with the organization's goals. Many organizations want to cut costs. They want to increase the bottom line. However, how are we going to achieve that while maintaining quality as well as maintaining the standards of the organization and really driving forward with this initiative?
Tom Stoneman:
Okay, so I'm going to be the CEO. So Mr. Harris, we're ready to invest. You sold us. Tell us what do we need to do to increase our readiness so that we actually utilize this and get our ROI?
Matt Harris:
So you've got companies that they're investing, but are they utilizing it to its full capability? So when we think about ready to invest in AI, we think about the components around the infrastructure, the models, the time and engineering effort that it's going to take in order to accomplish this. But have we considered what it's going to take from a data, a knowledge, a governance perspective in order to be successful with the investment that we're willing to make?
So I think a lot of times, I take a step back and I talk to them about their data. What are we trying to achieve and what data do we have available in order to drive the context around this agent that we're building or this RAG system that we're building or whatever the case may be? Do we have the right data? Is it in a structure that's able to be leveraged? And do we have the right context or the knowledge base built to drive this forward? Because if we want to make this large investment into this AI strategy, we have to make sure that we have a solid data strategy in order to drive that forward.
As well as when we take a look at that data, what sort of data do we have? What sort of compliance are we concerned about? How do we relate that then back into the overall strategy? Many times, organizations want to bypass the data strategy or bypass the data compliance or bolt on security at the end. Then what we end up with is we end up with what we consider a lot of tech debt. Tech debt in the industry is we have things that we didn't exactly take into consideration at first. So now, we've got to go back and re-engineer to accommodate that, which further elongates the return and further enhances the investment. So I think it's a lot about this planning stage before you get to that implementation stage.
Tom Stoneman:
The strategic argument is one thing, but selling it to the person who has to sign the check is another. Not all CFOs or board members are the same. I asked Matt how he handles the room when the budget holder is the last one to be convinced.
Matt Harris:
Tom, many times these days, especially with a lot of organizations that are working startup, venture-backed, they're getting a lot of pressure from the board to run down AI as quickly as humanly possible. So you've kind of got the enterprise level where the CFO's really concerned about spend and maybe is a little more hesitant about making that initial investment. To your point, what's the ROI? How do we achieve that? And then you've got more venture-backed startups that are running as fast as they humanly possibly can towards some sort of either AI-driven or AI-forward application.
So starting with more of the CFO centric that's more concerned about spend, they've got a budget that's already probably been approved. So with those sort of conversations, what I like to try and come to the table with is really speaking to those other executives or other champions for the change initiative and having that prepared for what does this look like for the organization? How is that ROI actually being calculated? And is it all tangible? Is it all dollars on the bottom line? Or is there some stretch goals in here that may be for driving revenue, driving experience, those sort of things that are not quite as measured on a P&L, however, they do enhance the overall organization.
Now, we have the venture-backed startups. They're running as fast as humanly possible to either get more capital, gather more investors, or do some sort of future acquisition where they get purchased. So those individuals, we have to sometimes take a step back and make sure that they're running towards the right thing with the right strategy because they see AI-driven, AI-native, and then we start using things like public endpoints for financial data or we start using databases that are not compliant, those sort of things that then cause a lot of issues for these startups later on down the road being less focused on being security-centric from the get-go.
Tom Stoneman:
When your work is abstract, dealing with data, strategy, and systems that don't exist yet, sometimes, the best reset is something physical. For Matt, it also involves horsepower.
Matt Harris:
It can definitely be a lot, especially when you're going through these long processes with large organizations. There's many layers of acquisition, so on and so forth, procurement, and that can really weigh on you. So a few things that I like to do that helps me blow off steam is over the last couple of years, I've really gotten into the gym. So working out a lot, I try to work out five days a week, take breaks throughout the day to walk a little bit. So that's really helped me, but my passion truly is I love cars. I've been a car guy for as long as I can possibly remember. So I love going to car shows. I have a classic late '90s Nissan Skyline that I imported from Japan, and getting out in that thing really helps me clear my head, gets me in the right space, and really helps me to blow off some steam. When I'm having those days where it's just a lot on my mind, a good drive always clears it up.
Tom Stoneman:
What's your favorite car show?
Matt Harris:
My favorite car show is probably either Slammedenuff or any sort of import expo. I'm a huge import guy. So I've always been partial to those sort of shows where you can see just some really awesome modified cars. Those are probably my favorites.
Tom Stoneman:
Oh, that sounds like a blast. I just recently was able to go to the Concours d'Elegance in Carmel. I don't know if you've been to that. That's a little bit of a different kind of show than I think the kind of show that you're talking about there. But I have had a few vintage cars, but I do absolutely nothing to restore them or anything. I just like to get them and drive them occasionally. So I guess you call me a lazy car aficionado.
Matt Harris:
Oh, I wish I just had more time, Tom, to focus on them.
Tom Stoneman:
Well, let's just blow off the rest of this and talk about that. Well, let's move into the next section here. I'm hoping you can walk us through an organization, you don't have to name names, great if you can, but you don't have to, but an organization that you worked with that really got it right and what they do differently from the start than maybe some other companies that haven't been quite as successful. And this is all about implementing AI, of course.
Matt Harris:
Absolutely, Tom. So an organization that I worked with very recently and continuing to partner with, and they were looking to really get AI-enabled, but they understood that there was going to be a process to this. They really wanted to get their data into a really good situation before they ever started down the path of AI. They had very distributed source systems of data that they wanted to be able to leverage. They had various issues and data quality, data governance, those sort of things were not in place at that point in time, but they really took that to heart that these things needed to happen in order for them to be successful, and they had a really good understanding of that.
We worked for several months with them, getting their data in a good situation, getting their data lake built out, transforming those sources into usable data, building out their analytics infrastructure. And then we started looking at what are the possibilities for AI? And we already even started doing that while building out their data lake. They had various sources that came in that schema-drifted every single month. So we were able to write AI agents inside of the data lake to help and support those issues as they progressed.
And now, we're at a point where we can really start to use a lot of this data to drive agentic processes. We're going to be able to add in an entire MCP layer around their various source systems to really focus on some agentic workflows. So they really took the approach of "We're going to get this right. We're not going to try to force bad data that will then allow for bad context into the business processes today." And by making that upfront investment and really understanding where they are and how they're going to get there, they have just really set themselves up for success for the future.
Any AI use cases at this point, they have the data pre-aggregated, they have everything they need to build good contacts, and we've already started deploying agentic analyst agents that are already supporting some of the core business units such as sales, finance, operations, and they're already being able to leverage those even before we get into the really cool agentic-workflow-type items. If they had not decided that they were going to approach this in a very structured manner, what they would've been dealing with is various distributed data sources that were living in multitudes of locations, which would mean that it would create a huge burden to orchestrate that data into some sort of model or agent. The data was extraordinarily dirty and it was sparse in certain locations, which would've mean you would've lost context. So some of these analyses, some of these agentic functions that they're looking to do today would not have been possible unless we had centralized this into a core location where we could actually do this.
Another huge risk, and I think this is the biggest risk of them all, is when you have siloed data sectors such as this, you have someone in finance looking at their data silo, you have someone in sales looking at their data silo, you have someone in operations looking at their data silo, then you're going to ask the question to three different people, get three different responses, and no one can trust the data. So you lose all trust and you lose all credibility straight out of the gate.
Tom Stoneman:
Tell us about, maybe give us an example of a company that didn't quite get it right. And at what point was it too late to do anything or were you able to interject something to fix and course correct?
Matt Harris:
Yeah, so there was a startup that I was working with a couple of years ago, and this was a case of a startup was going to be AI-centric. They were running down the path as quickly as they could to become that. And they were just, for the lack of better term, throwing it at the wall and seeing what stuck. So if there was an opportunity to embed AI in a core piece of the platform, it was going in it whether it needed it or not or whether it made sense.
So they were also to cut costs, because as a startup, you have a very tight budget, you're looking at investors, you're having to justify every expense that you have, they were using public APIs for sensitive financial information, which is obviously a no-no when it comes to any sort of compliance layer.
Then also on top of that, their database was not compliant with the regulations that they were needing to do. Then when looking underneath the code base, everything had been coded by developers that were not really developers. So you've heard the term vibe coding. There are good ways to vibe-code and there are ways that you may not should do that to put it into production. It's great for proof of concept or MVP, but once you get paying clients, there needs to be a little bit more rigor, a little bit more structure around your code base. And when looking under the hood, I was able to tell what was occurring.
So this organization decided that it was going to bypass any security protocol and just get to market as soon as humanly possible with a product that may leave itself and its clients up for a lot of risk. That organization, I consulted with them, we worked together for a little bit, and ultimately, we decided we were on two separate pages. They were really focused on getting to market. I was really focused on trying to bring security into the organization. There really wasn't much I could do to bring it back because it was a little too far gone by the time that I was brought in to support, and it just never really caught traction.
Tom Stoneman:
After years of these conversations, Matt has developed a clear sense of what's missing in most boardroom conversations. I asked him about the questions leaders aren't asking that they probably should be.
Matt Harris:
I would say the biggest thing that I would want to hear from an executive when it comes to the value proposition of AI is we talked a lot, Tom, already about we're going to automate what we already do today, but what do we own or what do we know that nobody else does? What do we have that no one else does? And then how can we further that conversation or further that thought into how we can be the most effective in our specific vertical with the knowledge that we have now enhanced with some sort of either AI agentic layer or some sort of bot, or how can we be even more successful because we have a market differentiator with this technology? Instead of just looking at the processes that are already existing today, what can we do that's even more than that?
Tom Stoneman:
And last question, Matt, what do you see coming that ... And particularly in enterprises, but it could be anywhere in the AI world, what do you see coming that's never been done before, some things that you think may happen that weren't possible even in three, four years ago?
Matt Harris:
It's a good question, Tom. So we've seen a lot of the software as a service or SaaS platforms get recreated and open-sourced over the last year. We've seen Adobe software, we've seen other entities get pretty well vibe-coded and then open-sourced that would've taken developers years to recreate, and that's why they haven't done it before. But as far as things that don't exist today, when you look at the layers of artificial intelligence, we're still very early on in our journey to reach any sort of generalized or super intelligence. We've got quite a ways to go in order to achieve those level of models where it's truly thought process, not just context-driven.
So as those sort of models come to fruition, realistically, the possibilities are truly endless if we have the infrastructure to support it, such as being able to truly go out and make purchases with literally just typing to a bot. That's totally possible. Being able to pick up items off of grocery stores, put them in your basket, walk out the door, and not have to walk through a scanner or not be looked at through AI cameras on a constant basis. Those sort of innovations as we get super intelligence with correct models are going to be very possible, but it also begs then the question of what security risks come along with that because as supercomputers get smarter, as these models are able to do more, that means they're also faster to break encryption, they're faster to be able to attack DDoS attacks. You could very well scale those out with the correct infrastructure and the correct model. So as we look at what's next, we've also got to prepare for what are the risks that come with what's next.
Tom Stoneman:
Well, thank you very much, Matt. I found your approach to all of these things very grounded and yet open to all the possibilities, and that's just a rare combination. So thank you so much for being on the show. Appreciate it.
Matt Harris:
Absolutely. Thank you, Tom.
Tom Stoneman:
There's a question Matt keeps coming back to in boardrooms and in our conversation. What do you know that nobody else does? It sounds simple, but many AI strategies miss it. They stop at cost reduction, efficiency, and doing the same things cheaper. In Matt's view, the organizations that go further are the ones asking what they can build that wasn't possible before and staying patient enough to lay the groundwork early.
Thank you for listening to The Intelligent Enterprise, a podcast where we get inside big ideas by getting outside of them. I've been your host, Tom Stoneman. Please remember to follow the podcast and leave a comment or review wherever you get your shows. See you next time.