Now, I probably wasn't the best person at like actually implementing those, but I did have a pretty good eye to identify technical solutions to business problems.

Host: How does Dualboot Partners empower organizations and deliver true business value through software development? And what are the core values that guide the company?

Ben: So, Dualboot has five core values, commitment, trust, community, teamwork, and a growth mindset. We use these core values to drive business value for our customers. If you think about what we sell, like what we sell is trust.

And so, we need our customers to trust us, to give us the flexibility so that we can deliver solutions that drive business value for them.

Host: How would you define application modernization? And in your view, how can AI be a game changer in modernizing legacy applications?

Ben: Yeah, so application modernization is pretty broad. It's basically applies to any type of software and really any kind of like almost technology that was built previously and is now at somehow obsolete for some reason or another. And that can be either the value that it's creating is it's not there anymore.

It's being overshadowed by something else. But in most cases, what happens is like the value is actually there that the software is generating, but some underlying core has changed. For example, the underlying tech stack there is old.

And so, there's no resources available to know how to operate on that tech stack. So, these are tech stacks like Fortran and Cobalt that are heavily used inside like banking systems inside of like older enterprise applications have all these. And the interesting thing here is that these applications are extremely successful.

They're very good at what they do. They generate a lot of business value, but we're getting to the point where no one can support them anymore. Another piece would be the obviously the introduction of like cloud technology.

So many of these applications run on premise, they run on mainframes, and like they don't they can't take advantage of the distributed cloud technology that's provided by the hyperscalers. So, in order to take advantage of the hyperscalers and this additional technology and to move these applications into a technical stack that can be supported by a vast number of resources that are available right now, we have to think about like, how do we move the technology? And it's interesting because we're not changing the value proposition, the value proposition stays the same.

And so, the question is, like, how do we take advantage of new technology? As we like dug into this, the realization that we had is like, in order to do this, all the requirements for this actually exist already, right? They exist in the source code of these legacy applications.

The challenge is, is like nobody knows how to get them, right? Because the folks who built these systems don't work at the companies anymore. They might not even be working at all anymore.

And so how do we extract this? I'll give you an anecdotal example. We had a bank came to us and they said, hey, we need to we need to modernize our legacy banking system.

I said, OK, great. They said, but we have some customizations. I was like, OK, well, how many customizations do you have in this banking system?

They said, well, it's on the order of magnitude of like seven to eight thousand different customizations were made. I was like, oh, OK, interesting. Like, what do these customizations do?

And they said, well, we don't know, like because the people who wrote them are all gone. And so, they know that the banking system works, but they don't actually know what any of the customizations that they do, what they specifically do. And that presents a political problem for the technical technology leadership within the company, because they essentially need to go back to the CEO board who's asked them to modernize these systems.

And they say, all right, we need two or three years and several million dollars to get the project started. And this board says, great. So, then we're going to be done with the modernization.

At that point, the technical leadership has to say, no, that's just to understand what our current system is doing so that we can now provide an estimate and a project plan for how we would actually do the modernization. So now you have a technical leadership who is supposed to be in charge of these systems, basically having to admit that they don't understand exactly what they do or how they work or have the ability really to maintain them. And so, this is where we see like AI could really step in, because what we need to do, we know that all those requirements exist in source code already right?

What we really need to be able to do is we need to be able to extract those requirements or reverse engineer them out of the source code. And so what we've done is we've built a tool called 3PO that helps us to do this process that basically accelerates the process of extracting or reverse engineering requirements out of source code so that then we can have a conversation with our clients around, OK, these are the things that the system is doing.

This is what you need to migrate. Maybe you don't need to migrate all of it. Maybe some of it goes to Salesforce or someone goes to SAP.

So, the other things that are proprietary, maybe get migrated into a new system. And then the actual migration process is a code generation process. But we have requirements now.

We have original source code. We really need to translate that into another language. And so now you can start to use tools like Cursor and Copilot in conjunction with your requirements system to start to generate code that would maybe take months or years to write by hand.

That code can now be generated in a matter of weeks or months when those tools are being supervised by expert operators. So that's a long way to say like we're seeing a lot of progress like with AI accelerating the ability to do these modernization projects. The other interesting thing is like it's not really a reduction of like technical capacity there because we're actually just taking projects that were otherwise financially non-viable projects and they're now becoming financially viable.

So, we're actually expanding like the total addressable market of the amount of software, amount of projects that really need to be executed on.

Host: Thank you, Ben. That was very insightful. And now let's talk about app modernization versus system modernization.

Like what are the key differences between system modernization and application modernization?

Ben: Yeah. So, the biggest difference is application is kind of what you think of. It's a SaaS tool usually or maybe a correct tool that you're using.

Whereas system modernization applies to a more broader perspective. So now you're talking about like the hyperscalers, like the ecosystem that you're actually operating in, security, kind of all the pieces that go around it. We focus, we do a lot of like application modernization is really kind of like the point that we come in at.

However, like there's typically a larger project wrapped around that, which is the system modernization that has usually like sparked the application modernization. Some, they had a directive that says like, hey, we need to move to the cloud or we need to move off of our mainframes. And then that, that modernization effort that sparks like, okay, in order to do that, we need to like modernize all these applications so that they're capable of running in these environments.

Host: Amazing. So, are there any critical considerations businesses must consider for like before they account for system modernization?

Ben: Yeah, there are several. The biggest thing is to think about like modernization as like a continuum, right? It's somewhere between, I need to replicate and migrate the value proposition that I'm delivering to my customers or my internal consumers, which gives you a lot of flexibility, right?

Which says like, Oh, I just need, I need to move the value that this is generating and that can be done by going to SAP or Salesforce or by like rewriting this in a different application with a different user interface. The other side of that continuum is like, I need to just replicate exactly what I have right now. And the really the difference between these two comes down to like, do you need to like maintain a common data structure?

Or is it using a data system that's being used by other systems? And so you can't just rewrite the whole application. You actually just need to like clone or replicate it, but you need to do it in a way that can be like, for example, containerized so that you can use the cloud providers to like deploy it onto.

So really thinking about the larger ecosystem of where you're going into will push you in the direction of like what your options are when you're doing these modernization projects.

Host: Let's shift to a major milestone. Dualboot partnered with AWS in June 2025 to fast-track cloud adoption for SMBs. So, what initiatives have been rolled out since then?

Ben: Yeah. So, we do a lot to help SMBs. There's a couple of different like angles that we take.

One angle is simply like the expertise that we bring to the table. So when you have a SMB, Austrian or Ideana, NavKit, kind of like someone getting started out, one of the things that they need to do is they need to have like technology team, right? If you're, if you're going to build a software application, not a technology application, you need to have that.

So, your options are you can build your own, you can use, or you can use somebody like us who like we have an experienced team. If you build your own, like that seems like a good idea on the surface, but typically SMBs missed several, several complexities in there that make this project much more difficult than they would expect. And the number one thing is actually just team building.

It's like, they don't have a team. So, they have to, step one is like, they have to build a team. And when you build a team, you have to form the team.

You have to norm the team, that team has to like storm, and then it has to be able to perform. And like those, each of those steps is not an easy thing to do for a team, right? And so, if you, if you work with somebody like us, we already have the team in place, these teams that executed on dozens and dozens of different projects.

We have hundreds of projects under the belt of the entire company. And so, we've seen the execution of this like many, many, many times. And so, what that gives us is it gives us an advantage because like, we know the teams, we know the tools, we know the process.

And so, we know all these things. And so, we take that off the table as a risk. So essentially like reducing risk for these S and B's and ensuring that they can then focus on their, essentially their go-to-market strategy.

Now, how does that play with AWS? So, one of the things that we do is like, as the hyperscalers continue to grow and get bigger, their ecosystems get much, much more complicated, right? So, if you just think about the number of tools that AWS has, like there, there's nobody who understands all of the tools, right?

And so, like now you're talking about, okay, we need to have domains of tools. And so one of the things that we have of several AWS certifications, like lots of our team members are AWS certified as well. And so, we have expertise like in these hyperscalers to help our clients understand which of the services that they offer, like would be useful for that.

Um, one example is that AWS has a service called Bedrock. That service can do a lot of things that most SMB's are trying to do with AI now. So SMB might come, if you form your own team, they might say like, Oh, team could say, Oh, we could build that.

And they start to like piece these things together from different sets of tools. And they're trying to pull it all together themselves. Whereas somebody like AWS has a service that potentially that offers basically all of those tooling that they need to make this stuff work.

Um, but they don't have the expertise on AWS that know how all that tooling works. And so if you come to somebody like us with like partners with AWS, like we know how all that tooling works and we can listen to your, the business problem that you're trying to solve and then make a recommendation around the types of tools that you should be using and by leveraging tools that are already built, like we can move that process much, much faster for our clients.

Host: Now let's talk AI. How do you see AI agents transforming day to day businesses, business tasks, and any use cases you'd like to share?

Ben: Yeah. So, we see this internally, like at our own company at Dualboot, we see this with our clients, so it's definitely happening. Um, it's, there's a lot of noise, right?

So, like separating the signal from the noise, is challenging. And I like to tell people it's easier for us to see like what it's going to look like five to 10 years from now than what it's going to look like five months from now. Right.

Cause we kind of all can see what it's going to be capable of, but how we get there is not exactly clear. So, the strategy that we take is like, we are looking at internal processes that we have, and then we provide this same expertise to our customers, we look at our customers internal processes, but in this case, we are eating our own dog food and like, we do this as well. So, we look at our internal processes and we say like, hey, where do we spend a lot of time?

Um, where do we spend that time? Where most of the work is being done is digital work, right? So, it's being done on the computer.

All that, all the information that needs to be done is on the computer and things like that. Um, and what is the level of effort or expertise required in order to like execute on this? Is this really complicated?

Like thinking work that requires like a lot of coordination between people, or is this like execution work that requires like typing on the keyboard, basically. Um, and so what we've done is we have a methodology and AI first methodology for the software development life cycle called DB 90. And what it does is it combines best in class tools.

Some of those are commercial tools. Some of them are tools that we built to like fill gaps. And what those tools do is that they provide custom interfaces for AI agents to execute on very specific tasks.

Um, so if you think about the life cycle that we have, we have our 3PO tool, which provides an interface for us to extract requirements from legacy software systems. So, okay, we can, we can use these AI tools to dramatically accelerate the pace at which we can extract requirements and get to like a backlog that says like, okay, this is the things that we need to do. Uh, did we use commercial tools like Figma, uh, to do, um, uh, designs, right.

And so, we are then feeding into Figma, like a set of screens that we need and a set of workflows and Figma can now start to take those and, and do a fat first pass at design. Does it get a complete? No.

And so, you still need like your designers, you still need your experts supervising those tools to make the changes to them as, as they, as they, uh, as they continue to develop. But those tools also get better over time. So, we're seeing accelerating, uh, productivity within those tools.

And then we use like Cursor, Copilot, and some of the, uh, code generation or pair programming tools, uh, and we, we use an MCP connection to connect back to 3PO so that we can have those tools extract and get access to the actual requirements for the underlying systems, which makes those tools, which are already pretty good, even more at. So now we're expecting our developers to supervise the code generation, right? So, you still need to be an expert in terms of like, what does the architecture look like?

What technology stacks are we using? What's the API structure look like? You like somebody still needs to like define all of those things, but the actual typing of the code can now be done by AI, which will, which dramatically increases like productivity.

And so, we start looking at like each phase of what we do, and we say, okay, what's the tool that we should be using here? How do we put an agent in place to take off some of the work that would otherwise be the long pole in the tent, but it's not necessarily the like very strategic work that has to be done. Um, and the goal here is basically to get 50% more productive, like in our process.

Like I think the lasting impact of what we're seeing with AI agents is going to be like an increase of productivity, right? And that's not, this isn't novel. This is like essentially the introduction of all net new technologies essentially leads to a productivity increase.

You go all the way back to like the introduction of the tractor and things like that, if what it results in is a productivity increase for the people that are operating in those fields, which then allows us, you know, more capacity to do other things. And I think we're at the very beginning of this and what we're seeing, what we have seen in the last 50 to 75 years is this introduction of like knowledge work. And now what we're seeing is the introduction of productivity enhancement or agents in this case for that knowledge work.

And so, it's probably going to follow a similar pattern to other like technical innovations. Now it's probably going to be faster because it's knowledge and it's digital work, and so there's not a physical constraint outside of electricity and data centers on how fast we can go. Um, so we're going to see some rapid change, but it's going to follow like largely similar pattern to the introduction of other new technologies.

Host: Absolutely. Speaking of scalable solutions, how does dual boot staff augmentation model work? Like what benefits does it offer to clients?

Ben: Yeah. So, what we can do with our staff augmentation model is we provide expertly trained resources into our clients. Um, typically we do this because our clients have, in this case, the clients would have an established team and an established process, so they don't need entire like technology team to come in and like take something over.

They just need additional resources to accelerate the development that they're already doing. And the value that you get with working with somebody like us is like, because everybody is who works for us as a dual boot team member, we're not like subcontracting this work out to anybody. We're not a recruiter.

We're not just finding people and pairing them with you. Like we, these people are dual boot team members. They're trained by dual boot.

They follow and have worked in our db90 process in terms of using AI and the methodologies that we have in place. Um, they've, they've had exposure to the projects that we've worked on. So multiple projects, and now you're able to like pull folks like that in your teams and you get essentially a much greater acceleration of productivity from, from these folks than you get otherwise, if you just went through a kind of a traditional recruiter method.

Host: And lastly, how have IT practices and services evolved over the past few years? And like, how can organizations and professionals stay relevant in this ever-changing space?

Ben: Yeah, I think the biggest change that we have seen and will continue to see is that the IT sector is becoming, I'll say less technical and more like business outcome focused. So, if you think about 50 years ago, right. It was highly, highly technical with the IT sector.

They literally kept the vacuum tubes like plugged in and running at the inside of mainframes and things like that. And what we're seeing is like, as we are abstracting away the complexity, right. And so, the complexity is being like now consolidated inside of the hyperscalers like AWS and GCP, the vast majority of the IT sector now is focused on like value delivery for, for customers.

And I think that's going to continue and you're sitting to see it already in the software development. So, as we're having agents able to like write code, now the folks who are supervising those agents need to have a deep understanding of the application that they're building and the value that that software is bringing, not just the ability to write JavaScript code, right. They have to actually demonstrate the value.

And I think we're going to see this like across the technology sector, where it's going to become a much more blended role where those resources are expected to fundamentally understand the products that they're delivering and that the value that those products deliver to their customers.

Host: Ben, thanks for keeping it real and making techtalk so engaging. Thank you for sharing your journey, expertise and future forward ideas with us. We wish you and Dualboot continued success in your mission to build smart, scalable software for modern businesses.

Once again, thank you so much.

Ben: Thank you for having me. This is great.

Host: Thank you everyone for tuning into this episode of ExtraMile by KnowledgeNile. I'm your host Sayali, signing off. Till then, stay tuned for more insightful episodes with industry leaders and experts.