Subscribe
Share
Share
Embed
The Value Creation Mindset explores the decisions successful leaders make to create real, lasting value for their customers, teams, and businesses.
Hosted by A.J. Singh, each episode features candid conversations with founders, CEOs, and builders who have been in the trenches, unpacking hard-earned lessons on leadership, technology, strategy, and execution. No hype. No shortcuts. Just clear thinking, first principles, and practical insight from people who have actually built things that work.
Pat Full Edit Ready for Descript
[00:00:00] Pat: Hello, my name's Patrick Weikle. I'm a technology strategist with Microsoft Corporation. I'm here today with my good friend and colleague, AJ Singh from Modularis. We're gonna be talking about his brand-new release of Modu- Modularis 19. AJ?
[00:00:15]
[00:00:21] AJ: Hi, Pat. It's really good to, uh, to, to, to see you again. Um, yeah, Modularis 19 is, um, our latest release.
[00:00:27] Um, we've been in the code gen space, uh, for, as you know, a long, long time. It's, uh, 26 years, and this November it'll, it'll be 27. So I think the... Now, a l- little bit of the origin story might be helpful. Um, a long time ago I was working as a consultant, uh, and I was the chief software architect at a couple of dotcom startups and, um, very interesting, very, very challenging work.
[00:00:50] Um, I was at, I was at one of these firms, and again, I was working as a consultant, and one morning my m- manager, let's just call him Neil, uh, walks in and he brings me three new engineers, and he says, "AJ, put these guys to work." And I turned to Neil and I said, um, "Neil, we don't need any more people. For everything that the CTO is looking for, we are well-staffed, things are in control, and, um, don't need any more bodies."
[00:01:19] And Neil turned to me and said, "AJ, put these guys to work. We need the billable hours." That didn't sit right with me. I ended up turning in my two-week notice the next day, um, and I exited. And, uh, this was in 2000. I started Modularis along with my co-founder, Olivier, in late 1999, and this was the impetus for us to go full, full, full bore.
[00:01:45] But that is a telling story because What bothered me about this is that this is not about maximizing billable hours, it's about efficiently creating and delivering value. It's about being efficient in terms of how you execute engineering projects, specifically software engineering projects. So, you know, we named the company Modularis, and that stands for modular information systems, uh, a- along, you know, a very simple idea.
[00:02:17] Software should, should be modular. And if you think about what that means, a module is a reusable component of something, right? So again, very simple engineering idea. Rather than solve the same problem over and over and over and over again, and then I guess bill somebody for it, I know it's very lucrative.
[00:02:38] But rather than do that, the, the idea was solve the problem once and for all, package that s- that solution in a reusable module or component, and then you can reuse it. And that's what we've spent the last three decades doing. Modularis 19 is our latest incarnation, and I'm excited to kinda tell you more about it.
[00:03:03] But it's following the same principles that we always have been.
[00:03:07] Pat: Yeah, it's been a real priv- privilege watching Modularis grow from, you know, a, what was essentially a startup back in the day to now, I believe, well over 80 or, or close to 80, uh, commercial software launches. It's pretty, it's, it's pretty impressive record, AJ, and I just wanna commend you on that.
[00:03:24] You know, most companies, most companies in this space, um, you know, are either, like, pure platform or pure consulting f- firm. And, you know, Modularis 19 doesn't sound like either. Um, can you kinda tell us a little bit more about what you actually built?
[00:03:38] AJ: Absolutely. I mean, we've been a long... We've been around long enough to try every business model.
[00:03:44] So we have done consulting, we tried a pure product play. But Because we're focused on the efficiency and quality of engineering. You know, it's like I remember in the olden days, Microsoft had a prescriptive architecture guide, which is great. It was a guide which prescribed how we should, how we should architect products.
[00:04:09] And these, and these guides are wonderful. Um, great diagrams, great instructions. The problem is that whenever a team tried to, tried to implement them, as soon as you added a little bit of time pressure, all those guidelines go out the window. The people know what they're doing, but there's time pressure.
[00:04:26] It's like, "We don't have time. Let's just build it and fix it later." So the reality is that in, in the commercial product space, and, and here I wanna differentiate software development from product development. Those are completely different things really. In the product development space, you know, when we came out with our first, uh, .NET version of our product and we were marketing it and selling it, we, we realized, oh my goodness, we have built an F-16 and we're trying to sell it to people who have never seen a plane fly.
[00:05:02] That didn't work, so we folded services back into it. And at the end of the day- Makes sense When you're building commercial products that have to be stable and have to be scalable and have to run profitably and have to be easy to service, right? When you're, when you're doing that, you need to have expertise.
[00:05:27] Mm-hmm. And now we will talk more about, about how AI is impacting things in a little bit, I, I, I'm sure. But the need for that core expertise, that architecture skill, the en- the engineering skill has never gone away and is more important now than ever before. So the, the Modularius 19 really brings together four core components.
[00:05:51] It's a, it's a sub- it's a subscription, um, and it's four core things. One, we have a vertically integrated set of runtime libraries. These are licensed royalty-free. These are maintained and updated by us. We, we, we, uh, we, uh, provide various forms of, uh, source code licensing, so there's no additional risk there associated with them.
[00:06:13] Wh- why do we provide runtimes? Uh, those libraries are pre-built, hardened, battle-tested modules. Why build a new module when you can use one that's already proven? So we have built that in a vertically integrated stack. Um, that's one component. The other component is what we call Modularius Architect.
[00:06:37] This is deterministic code generation, which is model-driven. The idea behind Modularius, and this is something that we've mastered, I'd like, I'd like to say we are probably about best in the world at this, at model-driven automation When you design or reverse engineer, or even these days infer a, a, a entity object model, Modularis Architect will then, will, will then generate that 80% of the code base for commercial product that is beneath the surface.
[00:07:12] That's the stuff that often goes wrong. That's what has to be built right for the overall solution to be stable and scalable and run efficiently and be easy to maintain, and we build it in a deterministic way. It's white, it's white box, it is customizable, it is your IP when it is generated. But that works with our, that works with, with our libraries to provide a complete modular foundation for your product, right?
[00:07:46] And it's not just a one-time use. This, this is designed to be used, you can generate multiple times in every iteration. It is designed to make not just the building, but the management and long-term maintenance and upgrading of, of, of a commercial product much, much easier. In, in fact, we've had customers who, um, we first helped them ship a product 20 years ago.
[00:08:08] That product has been modernized over the years con- con- continuously and in a low-risk fashion because we have provided continuous updates of our tooling and our libraries, and it is still stable and generating revenue ev- even today. So that-
[00:08:24] Pat: That's really fantastic ...
[00:08:26] AJ: I mean, that's the second piece, right?
[00:08:27] So- Mm-hmm ... between the libraries and, and, and the code that Modularis Architect generates, that, that's about 80% of your final code base. The last 20% is the surface level stuff, right? It's your core, um, API surface. It is your web applications and your mobile applications and your desktop a- applications, and today your agents In the past, before Modulars 19, we just generated well-architected starting points there, and then, um, our, our customers' developers would then build those assets by hand on top of this, uh, on top of this foundation.
[00:09:06] It still provided a huge, a hu- a huge amount of acceleration and risk, and risk reduction. With Modulars 19, we, we've embraced gen AI but in a pretty smart way, in that our AI assistant, along with, um, uh, uh, coding harnesses like a, like a Claude code and a, and a Codex, for instance. We have provided a curated set of, of markdowns and guardrails so that you can dramatically accelerate the construction of that last 20%, all of the stuff that customers actually see, how customers actually perceive and receive value from anything you might be building.
[00:09:50] That's now accelerated by AI code gen. But it's done in such a way that the, the, the amount of tokens you might need to consume, it's probably 90% less. The, the amount of risk you are undertaking is much, much less, 'cause you're asking it to do only 20%, not 100%, right? So is it, is, is it gonna be perfect? No, none of these things are.
[00:10:15] But the amount of code that needs to be reviewed by humans is drastically lower. All of the determinist- all of the deterministically g- generated code, honestly, you, you don't need to look at. It's all built exactly the same way. It's all maintained and supported by us really, and, um, that, that, that AI assistance piece is new to Modulars 19, and I think in many ways it's, it's, it's a bit of a game changer for us and for our customers.
[00:10:46] But the fourth... Uh, go, uh, go ahead, Seth. You had a question?
[00:10:49] Pat: No, no, no. Please go, please continue. Please, please. Sorry.
[00:10:51] AJ: The, the, the fourth, the fourth piece is, as I mentioned earlier, expertise. We have tier one support baked into the model at our subscription plan. So- There is no case where you can be successful without access to real deep expertise.
[00:11:10] Do you need to hire that yourself? Well, you certainly can. Can you outsource it? Yes, you can. But a- as you know very well, Pat, being an architect for so many years, you know, we, we've both had customers call us after things are on fire and ask our advice. A- And in five minutes we gave them a piece of advice which would've saved them six months and $5 million.
[00:11:40] Having the right advice at the right time i- is really invaluable, and that's why baked into our subscription plan is unlimited access to our best architects and en- and en- and engineers. So those, those four components, we feel take a huge amount of the risk out of building commercial software products.
[00:12:02] And that's, uh, you know, we, we have 79 down and are looking forward to, to more.
[00:12:06] Pat: Yeah, me too. That's really, that's really great, AJ, and, you know, I think, I think the approach makes a ton of sense. In fact, you know, you mentioned, um, like AI code generation, et cetera, and, you know, what would you say to a, a CTO or a VP of engineering, um, that's sort of already made investments in AI code generation, but maybe isn't, isn't seeing the kind of value that, that you're claiming?
[00:12:26] W- What would, what would you suggest, um, is the difference between sort of what you're pos- positing and- Yes ... you know, just sort of vibe coding or, uh, whatever, whatever the, the, the flavor of the day is?
[00:12:39] AJ: Excellent question, Pat. The, the, like, vibe coding is great for I, I would say the bottom third of the market.
[00:12:47] If y- Mm-hmm ... if your, if your, if your solution really doesn't need to be very stable and scalable and profitable and serviceable, if it's just to solve a problem, wonderful. I mean- Mm ... it kind of fills the same space that low-code and no-code platforms have for years, for years and years. The, the way to think about it is that, and I don't know why it's, it's just kind of permeated now, but Anyone who's been writing code, and I've been doing it since I was 10, uh- Go
[00:13:19] go figure. Um, you realize pretty quick that more code doesn't necessarily mean more value.
[00:13:26] Pat: Mm-hmm.
[00:13:26] AJ: And the inherent nature of these LLMs is... Well, I, I don't know. Maybe I'm an LLM because I am notorious for long answers to short questions. And-
[00:13:39] Pat: That's what makes you an architect, my friend.
[00:13:42] AJ: Well, that's not what great and elegant code looks like.
[00:13:46] And AI's great at generating lots and lots of code. That does not equate to value. So what I would... The guidance that I would give to a CTO or VP of, of eng- of engineering who i- is s- spending a lot on tokens and time but struggling to actually deliver value, is understand that the timeless principles behind value creation and value delivery are timeless for a reason, you know?
[00:14:15] It might seem a bit trite, but it is garbage in and garbage out. That has not changed. Uh, if, if you're in the commercial or in the enterprise space, you can't sacrifice stability. If your product isn't stable, you, you have nothing to sell. If it's stable but it's not scalable, you can't hit any of your growth, growth, uh, targets.
[00:14:34] If it's scale- if it's stable and it's scalable but you can't run it profitably because you've used every wonderful service provided by either Azure or AWS, 'cause it's so nice and easy, if you can't make a buck, why'd you start? And fourth, if you've done all that and it's stable and it's scalable and it's profitable-
[00:14:52] Pat: Mm-hmm
[00:14:52] AJ: if you can't maintain it, you can't service it without sacrificing stability, scalability, or profitability- You might have solved the problem for a quarter, but when the customers want more value, order over quarter, you can't service them, and at best you will stall out and likely possibly worse. So those things haven't have, haven't changed.
[00:15:15] Use those as you're measuring SIX, regardless of how you're building it, right? That, those are, those are timeless things. When it comes to value... When, when it comes to delivering value to a business and out to the final end customer, honestly, that's still what gets me more excited than anything else, is these are all means to an end.
[00:15:36] The end is continuous value creation and delivery in ways that the business can measure. So great, you are in charge of building a product. How is that product helping you retain existing revenue streams? How is it helping you light up net new revenue? Have, have you actually measured it, right? Mm-hmm.
[00:15:55] It's, it... Just because you ship or it's in production, you're not anywhere close to being done yet, right? Um, have you, are you measuring a positive impact or negative impact on customer lifetime total value? When it, when it comes to the valuation on exit, nothing matters more than CLTV. So are you paying attention to it?
[00:16:16] And finally, uh, a simple thing, are you actually reducing costs through this investment in R&D? So if you focus on the basics, you'll, you will have a much, uh, a much better chance of actually creating and delivering value, and it's, it's the... On the AI front, th- th- that's the challenge. It's, it's driving so much change, and it's having people question so much, "Uh, all right.
[00:16:45] Do I know what I'm doing? What am I doing? Does it matter? Is it really 1,000 experts in a box? Oh, you don't need expertise a- a- anymore." All that's crap. You need expertise now more than ever. The under- the u- underlying rules of physics have not changed, and they will not change. So remember those, re- re- remember those, uh, fundamentals and measure the results and you'll, you, you will find the right path forward.
[00:17:12] That's what I would say.
[00:17:13] Pat: Yeah. I, I, I think that's amazing, and I'm gonna reinterpret, um, something that you said there because I think it's im- it's important, um, to call out. You know, a lot of what you're bringing, uh, through Modularis 19 as, as well as your own experience, you know, are really enterprise class philosophies, right?
[00:17:29] Um, but if I were to reinterpret something you said here, and that is that, you know, AI gener- c- AI code generation really, um- creates or accelerates a lot of technical debt, um, and doesn't necessarily, uh, that doesn't necessarily correlate to revenue. So, you know, if I were a, again, technology officer or even CEO of a, you know, small, mid-s- mid-size software company, um, break that down to me.
[00:17:55] Like, help me understand, uh, what that means and, and, you know, again, how, how I could leverage the, the platform without making a lot of the same mistakes or that frankly I can't afford if I'm, if I'm in that, uh, in that market space.
[00:18:09] AJ: No, I, I get you. I mean, you know, the, the Technical debt is what we've been calling this for a while.
[00:18:17] Maybe the new word is slop, the old word was spaghetti code. You, you don't need gen AI to create technical debt, and having some technical debt is okay. It's not something you should try to eliminate totally. It's not possible. It has to be managed well. And the, the challenge with kind of pure gen AI is, is you, you have to understand something.
[00:18:43] Gen AI builds products from the outside in. It's, it's, it's excellent at creating, uh, user experiences. It's, it's excellent at, uh, creating test cases that will always pass because they're coded to pass. Does that necessarily mean that they're actually testing what they should? Don't really know. But where, where technical debt really, really bites you, I mean, it's the...
[00:19:10] Uh, let me give you a, a visual. Uh, technical debt is the part of the iceberg that's beneath the surface.
[00:19:16] Pat: Mm-hmm.
[00:19:16] AJ: It slows down everything you are, you are doing. You need to make a modification, you need to add a new capability, you need to build a, a, a new system of engagement on top of a, on, on, on, on top of a system of a record.
[00:19:29] That is all drastically slowed down by this stuff beneath the surface. And by default, AI code gen will, again, it's, it's, it gives long answers to short questions. It's very good at producing a huge amount of content, and just having a human in the loop isn't really good enough because when you've got your best engineers simply re-viewing vast amounts of code, you will very quickly get to a point, and this is a human frailty, where they'll say, "Looks good."
[00:20:04] Looks good doesn't mean is good. And so the, the, the way that we've tackled this, and this is really where I think this is like a chocolate and peanut butter sort of thing Model-driven generation, deterministic code gen builds products from the inside out. Mm-hmm. From the model, the underlying structures out, and it gets you, again, up to about ei- 80%.
[00:20:32] If you can take that to the bank, that's a solid foundation. You're not building on a floating chunk of ice. You're not building on sand. You have a solid foundation. That couples really well with the outside-in approach of, uh, of a, of, of a generative AI. And that, that's, that's I think a path to producing the most value with the least amount of risk.
[00:20:54] And at the e- at the e- at the e- at the end of the day, you know, it's, it's, it's not... The person who wins isn't the one who has the most lines of code. The person who wins is who can create the most value the most efficiently and monetize that effectively. And by taking the risk out of the core, ultimately you are producing a lot less technical debt, and even the target, even the AI-generated code you are, you are, you are building out is only 20%.
[00:21:24] That means your token spend is only 20%, even less than the rest. So it's, it's just about engineering efficiency, which again, I, I, I, uh, I could go on and on, but I'm gonna
[00:21:36] Pat: stop. No, it's, it's really exciting. Um, you know, I'm gonna pivot back quickly to, uh, another point you made earlier. You really talked about, uh, essentially two, two sides or two engines, uh, of Modularis 19, essentially one that sort of creates the code and the other, other that just sort of ships the code.
[00:21:54] Um, how does the platform sort of determine which is which? Um, and why, why does that matter so much?
[00:22:04] AJ: Well, as Stephen Covey said, you know, "Begin with the end in mind." The end is building a revenue generation engine. When you're-- Whether you are building a enterprise app, used intra-enterprise, or you are building a commercial product which, which you are going to be selling, the ultimate job is to build a revenue engine.
[00:22:31] And I think what I mean by this, you have to ship and support a commercial product that is, again, has all the qualities needed to actually do that. Just getting excited about building something. I mean, everybody loves launching things, right? I mean, I watch, uh, well, I, I guess the Falcon line, uh, uh, Falcon 9 launches are pretty frequent now.
[00:22:53] But watching the Artemis launch, watch... It's, it's like you s- you see something like that and you see human progress. And yet, you know, when, when, when the rocket's taken off, the rocket engineers are always thinking about the billion ways this thing is just gonna blow up. And you need those experts thinking in those ways to prevent that from actually happening be- because you gotta get the payload to orbit.
[00:23:17] If you can't get it to orbit, all of the effort is not worth much. So I, I... Or, or you can actually end up creating negative value. So I s- I suppose this, thi- this, this is why our, uh, our perspective is so much focused on production launch, and not just a one-time launch, but continuous delivery of value and new features, new capabilities that, that is necessary for you to keep that, keep that payload in orbit and generating revenue.
[00:23:50] So-
[00:23:51] Pat: Yeah, that makes, that makes a ton of sense. That makes a ton of sense. I, I'm glad you framed it that way.
[00:23:55] AJ: Yeah, I, I... And, and, and so all, look- At the end of the day, what Modularis 19 is about, what, what Modularis is about, what our product is about, is it's just a distillation of all of the experiences that, that, that our architects and engineers have, have gathered over the last three decades, and we poured this knowledge into the product.
[00:24:19] It's providing guidance only... I mean, again, the guidance will get thrown out when you apply pressure. The only way to really deliver it is through automation and packaging. That's what Modularis nine- 19 is. It is that distilled intelligence, intellect, experience packaged together through the right kind of automation, and again, also backed up by the people who learned it, who, who learned things the hard way so they can support you and guide, and guide you.
[00:24:50] That's what we've, uh, we believe that that's the right way of creating and delivering the most value to our, to our customers. And as our new T-shirts say, right, "Modularis, generate value, not slop." So that's- I love
[00:25:04] Pat: that. I love
[00:25:05] AJ: that ... that's what we're about.
[00:25:06] Pat: I'm, I'm still waiting on mine, by the way.
[00:25:07] AJ: Uh,
[00:25:07] Pat: it's coming.
[00:25:08] Hey, so, you know, you mentioned earlier the name Modularis is, is, you know, really means modular information systems. I think that's, I think that's actually, you know, pretty slick. Um, and you also sort of mentioned that, um, you know, the, the name, the, the approach is sort of how the name developed, and obviously how- Yes
[00:25:24] how in practice that really helps software companies sort of accelerate their development, uh, to ship the product, get, and get a revenue stream going. Um, in fact, if, if I recall, I think you've made some fairly bold statements around, you know, being able to get to a product release, uh, in, uh, forgive me if I quote this wrong, but 90 days.
[00:25:44] Um, uh, you know, and I think about, I think about teams that have, you know, um, smirked at that, at that, um, uh, you know, that proclamation, especially when they've struggled to, to reach that kind of goal in the past. Uh, what would, uh... What do you say? How do you go about delivering something like that and proving the value that quickly?
[00:26:00] AJ: Yeah, no, I, I, I, um, that's- That's been something that, um, we've spent a lot of time on. I
[00:26:07] Pat: imagine so.
[00:26:09] AJ: To, to... And to really deliver in 90 days, it takes more than just automation technology, code generation, or even architectural expertise. At the end of the day, you have to do a decent job of helping the customer do the discovery and design
[00:26:28] Pat: work.
[00:26:29] AJ: How do you... Uh, again, what defines success? It's gonna be different in each domain, for e- for, for, for each customer. So oftentimes, I mean, when we first engage, we spend a couple of days working deeply with a, with a customer in what we call a product launch workshop. They have ideas, they have concepts, they might have, um, a, a vibe-coded POC.
[00:26:57] That's all wonderful stuff, right? Mm-hmm. If you capture the, the vision and define the scorecard for actual launch, if you don't have a scorecard to know if you're successful, what you get to in 90 days won't really matter. So we s- so we start with that, and then we, again... I mean, this is not our first or even 79th rodeo.
[00:27:20] We're now up to our 80th rodeo, and this is a combined, uh, integrated offering where you have got access to our experts, our architects, and you have access to our technology. We bring that together to- Mm-hmm ... to, to take the output of, say, a one to two-week dis- di- um, discovery and design effort, and then automate as much of the product development life cycle as we possibly can using a mix of deterministic and, and, um, LM-driven code gen with experienced architects guiding the process.
[00:28:00] And 90 days is, is, uh, i- is a mark that we hit, uh, couple years ago. Honestly, with Modularis 19, I think we can get that down to 45.
[00:28:11] Pat: Wow. That's pretty bold. That's a pretty bold statement.
[00:28:14] AJ: Well, I mean-
[00:28:14] Pat: You heard it here first, folks.
[00:28:18] AJ: We're working on it. Yeah. And the, the, the... If we can help customers light up net new revenue in 90 days or 45 days, uh, it's really the only path everyone wins, right?
[00:28:30] I mean, that's, that's ultimately... It's, as they say, sales solves everything. We understand-
[00:28:37] Pat: I love that. I love that. So AJ, you know, talking to CTOs and CEOs, you know, as I do, um, they've often assumed that bringing into their, uh, their organization Uh, was somehow gonna, you know, resolve all of their technical debt or, or, you know, basically h- give them some magic wand to overhaul their legacy code bases, et cetera.
[00:29:01] Basically, you know, we're finding or, you know, what I see is that while AI can help, um, it doesn't necessarily unlock them the way that they would expect. Um, can you tell me about a time when, you know, uh, I would say Modularius or the platform itself helped them really realize that, that vision of being unstuck, uh, in their, in their code base and, and sort of paying down that technical debt at a rapid pace?
[00:29:24] AJ: You know, one of our customers who we've been working with since, uh, since, uh, 2013, uh, I was on a regularly scheduled, uh, monthly sync up with, with one of their managing directors, and he, uh, brought up an interesting challenge. Now, they have hundreds of on-prem deployments of their flagship ag product, and they, they were getting hit up by their customers who were asking for agentic access to their underlying data.
[00:29:54] They wanted to query and analyze underlying data. They wanted to automate, uh, particular workflows using GenAI. I mean, uh, a perfectly logical ask. And he asked me, "AJ, can we enable that sort of automation? Can we enable that, that sort of agentic access, do it in a secure way, um, without rewriting everything, uh, you know, using the assets we already have?"
[00:30:19] And the answer is a definitive yes, and in fact, we are in the middle of, of, uh, of, um, uh, of actually doing this now. Uh, this is making full use of, um, three core capabilities of, uh, Modularius 19. One, our metadata-driven MCP server, which is wi- wi- which is built in. Two, our, our agent gateway service, which has already connected all of these on-prem deployments with a secure cloud-based server, uh, uh, a secure cloud-based, uh, gateway, uh, rather, that, uh, allows secure bi-directional access to that on-prem data.
[00:30:58] And then three, um, a virtual model, uh, of their, of their underlying domain. Now, the virtual... Now, the, now the virtual modeling capability of Modularius 19 is truly unique. We've invested very, very heavily in this. What this lets you do is actually come up with a virtual model of, of the underlying data which exists in an existing legacy product So it compensates for any deficiencies in, in a legacy schema.
[00:31:29] It looks the way you would want a brand-new version of your product to actually work from a schema standpoint. But here's the amazing thing, it runs on the existing legacy system without any breaking changes. So it's the best of both worlds really, and we have combined this virtual model, the agent gateway capability, and the MCP server, um, to enable, uh, uh, customer service representatives, and in some cases even an- e-even end customers to securely access their on-prem data in an intelligent and an efficient way using the AI, uh, orchestration platform of their choice, whether it's Copilot, Claude, or Gemini, it doesn't really matter.
[00:32:20] So I think this is a really solid win. Uh, the, the return on invested capital is amazing because we are reusing all of these existing assets, and it's certainly possible to light up a whole net new revenue stream for our customer. So I think that's the one clear example where I, I, I think, um... or, or I think that's one really clear example where we're able to help customers create a ton of value, um, with very little rework, and deploy AI in a way that truly creates value rather than just drives increased consumption.
[00:33:02] Pat: Well, hey, AJ, I could not be happier and more excited for the launch of Modularius 19. Personally, I can tell you, I am so proud of you guys. Like I said, I've, I've watched you grow this platform and grow as a company over, well over 20 years now, and I am just thrilled to, to talk to you about it today and really appreciate your time.
[00:33:25] AJ: Thanks so much, Pat. I, I, uh, really appreciate this. Coming from you, it really means a lot. Thanks.
[00:33:32] Pat: Well, likewise, my friend. Congratulations.
[00:33:34] AJ: Thanks for listening to this episode of The Value Creation Mindset. I'm AJ Singh. If this episode helped you, please follow the show, leave a rating, and share it with one person who you think it will benefit.
[00:33:47] For show notes and more, visit modularius.com/podcast. See you next time.