#411 Moving Fast Without Breaking Everything: MVP Strategy with Andrew Duncan

[00:00:00]

Mehmet: Hello and welcome back to a new episode of the CTO show with Mehmet. Today I'm very pleased joining me, Andrew Duncan. Andrew, the way I love to do it is I keep it to my guests to introduce themselves. So a little bit about maybe [00:01:00] yourself, your experience and what you're currently up to, and then we can start to take the conversation from there.

Mehmet: So the floor is yours.

Andrew: Yeah, absolutely. Well, thanks for having me, Mehmet. I appreciate it. Uh, I'm Andrew Duncan. I'm the founder and CEO of Vertice Labs, and we specialize in boutique, you know, software engineering for startups and small to medium sized businesses. And, you know, I think what we bring to the market that's unique is, um, in my experience, you know, I've worked in industry and consulting for almost two decades, and there's a lot of bloat and, uh, misalignment in the software.

Andrew: Process building products or internal processes. And so we just cut through all that. We aligned strategic, uh, management and boots on the ground teams to deliver effectively and efficiently.

Mehmet: That's great. And again, thank you, Andrew, for being with me here today. It's, it's a topic that, although, you know, As much as you know, I I keep discussing it.

Mehmet: I always keep, you know, discovering some new [00:02:00] insights And this is why I was happy, you know when we get connected and you know We decided to do this episode together because I believe you know, we're not discussing this Much enough, in my opinion, because the idea here, uh, Andrew and, you know, like, maybe this would be my first question.

Mehmet: Um, so especially when we talk about maybe startups, right? And founders who might not also have a kind of the technical capabilities of developing their own software. So. Why it's good to come to someone like, you know, you, you and your company to get things done the proper way rather than, you know, just keep trying and, you know, paying a lot of money.

Mehmet: So, you know, let's enlighten this area a little bit together.

Andrew: Yeah. And that's a great question. Um, so I'm going to short, like if you're not technical founder, you know, and you have a venture idea, [00:03:00] um, Um, And you want to get to market quickly, you know, hiring an expert like Vertice Labs is a surefire way to get it done, get it done right, and then get it done well, quickly, right?

Andrew: Um, so, you know, we offer guarantees and support as a standard part of our, our offering when we engage with clients. Um, but I also often tell founders, you know, you, you, if you're building a tech company, you do want to hire a CTO at some point, right? Um, sometimes you just don't have the time. To, to find that right person if you don't know someone already in your network.

Andrew: So we're a great option for that. We can get you off the ground, get you launched, and we can even help you find that CTO. Um, I've seen, you know, I've seen time and time again, founders, you know, start, they'll hire, you know, a, you know, a contractor and get started, or, you know, an offshore firm. Um, and those can get, Sideways pretty quickly because if you don't have that strategic technical leadership guiding the the engineers or the team or the [00:04:00] offshore team, um, it gets sideways, right?

Andrew: They can paint themselves into a corner, pick tech stacks that, uh, or architectures that don't quite scale. There's all kinds of obstacles that you've got to navigate, you know, yeah. If you're launching a mobile app, for example, their gotchas through the approval process, you know, there's if you're in the fintech space, there's compliance that you've got to worry about.

Andrew: So, you know, we, we bring that, you know, knowledge already baked into whatever we're doing. And so you're not, you're not tripping over those along the way, which ultimately extends the time to launch and extends the cost as you're paying for the resources.

Mehmet: Fantastic. Now. I would like a little bit to spend some time on the startup and especially, you know, maybe, um, you know, for, for the folks who would be at the verge, I mean, let me first clarify this.

Mehmet: So do you help also in, you know, the MVP stage, Andrew, like to, to get like, at least kind of a, Working first version. Do you have? [00:05:00] Yes,

Andrew: absolutely. So we, we, we take it from, you know, I, you have an idea you walk into or hop on a zoom call these days, uh, and you have this idea and it can be really abstract.

Andrew: You can have this abstract idea. Uh, and we help you, you know, whittle that down to, you know, a specific niche and a problem that you want to solve, right? Because like building software for the, for software's sake is inexpensive to no end. And what you really want to do is focus on a customer problem, a pain point that you're solving.

Andrew: And so we help founders, you know, identify the specific pain point that they want to solve. And we're really big on, Being niche, being specific, you know, so you don't want to boil the ocean. Um, you don't even want to try to boil a pot when you're building the first version of your MVP. You want to solve a very specific problem, get that out, test the market, get some people on the platform to use it, and then take that feedback to drive the next, the next iteration of your technical implementations.

Mehmet: Right. So, so they, so [00:06:00] we start from, from, from the MVP, uh, Andrew, but at some stage, you know, once things start to work properly, let's say, of course they need to go through different iterations until maybe they can find the proper also product market fit. Maybe they have to a little bit shift things down the road.

Mehmet: Now, And I'm taking it from startups and then I will, you know, shift to the next phase. So do you advise them to have the team? I mean, the development team as in house or you still tell them, okay, as leadership consultancy for software development, like I gonna bring, or like, maybe I would help you bring someone part time.

Mehmet: And once things You know, start to pick up. Then we're going to make this team, you know, full time with you and then we will take you to the next stage. So walk me through this process. Like when, when this shift happens and when it applies, uh, [00:07:00] across each, each, you know, phase of this growth of the startup.

Andrew: Yeah. So that's a great, great question. So when you first, when you're first starting, You know, we can bring the technical team, right? So we, we, we do, you know, turnkey ideation to launch, right? Uh, software development, product development, um, and so we can bring that if you have a team in house already, right?

Andrew: As if you're further along that curve, we can help you align the processes, the strategic vision with the boots on the ground, um, and make the process efficiency efficient. So, you know, there's a bunch of it. You know, technical aspects of the software development life cycle, uh, that you can build in efficiencies into.

Andrew: And so we help teams and companies navigate that. So I have a couple of clients now that, uh, they have, we have built a team, right? And they're using Ortiz Labs. Technical talent team to deliver their MVPs MLPs, which I like to use instead of an MVP, which stands for minimal lovable product Right, you want to build something that you actually can get valuable feedback from um and the [00:08:00] ultimate goal is to ramp up a technical team internally because they're a Technical product they're software as a service platform.

Andrew: Um, and so As time goes on, they, they're going to launch the MVP, and that's launch is scheduled for March for this particular example, uh, then we are going to stay on for some time to mitigate any bugs that might creep up, uh, any feature enhancements, because they will get early feedback from, from users, uh, and then, uh, We will phase in, uh, their own team that they're going to hire.

Andrew: And I'm going to help them out with that. Right. Uh, so such that they can have their team that is immersed in their business. They talk to day in and day out. Uh, they're engaged with the clients and, and, you know, understanding that client feedback and, uh, Ultimately, we're we're we'll get down from that project.

Andrew: And then I have another client who is we're doing the bill for them just the same. And the first person we're gonna help them find is their CTO slash co founder after we launched their MVP or actually kind of along the lines of [00:09:00] while we're building it. And that that's an example of where the founder wanted to move really fast.

Andrew: The founder has some technical capabilities. Uh, and so they want to move really fast. They have a you know, a seasonal product they want to launch by a specific time to capture Uh the season so

Mehmet: yeah fantastic now on on that, uh, you know, uh, you mentioned something very important which is related to aligning also The technical team wears, you know, the, the biggest picture, right?

Mehmet: So, so everyone needs to be on the same page. So what's the best approach to do this, Andrew? Like, and, you know, this is why I love, you know what, this is why the called CTO show, right? So the CTO is the, is the guy who's main responsibility to align tech and, uh, and business. So, uh, what's the best approach that you have seen working for making this alignment?

Mehmet: Perfect. [00:10:00]

Andrew: Yeah, that's a great question. Um, so I was a CTO for many years in the industry. Um, and so what I found is this sort of a, a unique skillset of, you know, sitting in the boardroom and, and with founders and discussing where the company needs to go, right. Long term. Uh, and how to get there. And then sitting in the, the individual teams, software engineering teams, and talking about what we're doing tomorrow, right?

Andrew: There's a, a bridge that you, a gap that you have to bridge to do that. And so, what I have found is, and I say, I talk about this all the time with my clients, is You, CEO, COO, CFO, you guys have to continue to stay focused on the long term and what we want to do 12, 24, 36 months from now, and then have those conversations and have them in depth and detail.

Andrew: And then my job or the CTO's job, whoever's sitting in that role, their job is to, participate in those conversations, talk about what is capable, [00:11:00] what we're going to need to make the, you know, what's capable happen. And then they have to turn around and go sit with the technical teams, with directors and team leads, and talk about what we're doing right now, right, at a detailed level enough that people can write code for it, right.

Andrew: And, but also communicate that in not too distant, right, not the too distant future, we're going to be doing this. And so they provide the context of what's happening right now within the larger vision. Um, and, and an important aspect of that is, is, um. is focusing on what the businesses, businesses goals are.

Andrew: So like long term goals and short term goals, what we're working on right now, we have a very specific intention, um, and desired outcome that we want to achieve. And so aligning that specific outcome with what the actions are that we're taking today with the context of what is in the future is the skill set.

Mehmet: Right now, this is very [00:12:00] insightful, Andrew, and I think it's not. It's not an easy task. It's like a big responsibility also as well. And now not only as startups, even like in, in small medium businesses and even in the, you know, big, big enterprises as well. So one of the things that, you know, as leadership.

Mehmet: you know, you need to make sure is you are not wasting your resources. And when I say you're not wasting your resources, that's both as human resources or kind of, you want to call it like a wasting the time of your team and of course, wasting money. So from your experience and you know, maybe you can tell me more about the role of, you know, utilizing like consultancies, consultancies, like.

Mehmet: You know, the one you have, Andrew, for effectively managing, you know, uh, not wasting people's time and also, of course, not, not wasting the [00:13:00] money as well.

Andrew: Yeah, 100%. So, you know, in software engineering, you have your teams, your engineer teams, and you know, you're paying them salaries. Right. And so if you waste their time, you are wasting the company's resources.

Andrew: And in this case, it's capital resources and opportunity, opportunity costs. Uh, and this is something I'm super passionate about and part of how. I think about when I'm building what I like to call high performing software engineering teams. Um, a huge aspect of that is kind of what we just talked about previously.

Andrew: That is making sure that whatever we're focusing on has a desired outcome and keeping that desired outcome as the lens through which we look on for the actions that we're taking right now, right? Um, and so, you know, aligning every effort with some, some desired outcome that is clear, that is crystal clear that we know.

Andrew: And can measure what that outcome is, is a big piece. That's sort of step one. And then step two, a big thing that I, I talk about is, um, anytime a development team is idle, you're just burning cash, right? [00:14:00] And so I talk a lot about keeping the team fed, and that is making sure they have the next highest weighted priority.

Andrew: Tasks to work on. Right. So whatever the effort is, it's, it's prioritized by the business again, with the context of that desired outcome. And it's clear enough that an engineering team can go execute on it because computers only do exactly what you tell them to do. And so if there's an ambiguity, then The engineering team will have ambiguity.

Andrew: They will either make assumptions. You know, I'd say they make thousands of micro decisions every day. So they'll make assumptions, they'll write the assumptions into the code, and that comes out on the other side as missing the target of the desired outcome. So being really clear about what we're doing so that we can hit the target, and then keeping the context of what that target is, are two really key aspects of that.

Andrew: And that happens through many conversations between the executive team, and And the director level for larger organizations or just you know founders and the development team and startups,

Mehmet: right? So [00:15:00] talking about also avoiding failures, um and it the context i'm aware that it's different between like maybe a small or organization startup versus big Big companies.

Mehmet: Uh, things can go wrong. So in startups, for example, you know, we always hear about the term and that we discussed it many times on the show also as well, which is, you know, like after a while, you discover, you start to have a technical depth and this technical depth can Um, you know, it results off something to fail.

Mehmet: And then you figure out all like what all what we were doing all this time again back to the waste wastages, right? So So it's like kind of wasted. And in big organizations, we see, okay, they have this initiative. They start to work on it. So is it like, is it like lack of understanding the technology? Is it because we are trying to move too fast?

Mehmet: I'm trying to first understand, you know, the, [00:16:00] the root cause of the problem and then, you know, how I can avoid, you know, Uh, failures by maybe putting some guidelines or whatever you want to call them. So I make sure, of course, failure is possible. That's everyone can do mistakes, but I mean, reduce the risk of failing as much as possible.

Andrew: Yes. Yes. Great question. So, you know, Startups can fail for a number of reasons, right? And one of which could be your product doesn't launch or your product fails when it launches and then there's a whole host of other ones like you don't have a product market fit or Um, you've gotten a bad reputation in the market because your product doesn't work, right?

Andrew: And so that's a technical debt failure or you don't have the right sales or marketing funnel or you're not solving the right problem, right? So there's a number of reasons but like just to kind of hone in on How it can fail with just the technical aspect. It's in in in and of itself a quite complex problem.

Andrew: So a lot of it has to do with kind of we just previously talked about around [00:17:00] aligning the vision with the tactical right and so that's a that's a piece that often. Often gets missed, right? Just sort of translating, uh, the goal of the company to the technical team and having what they produce to align to that goal, big piece of it.

Andrew: Um, and then the other pieces I would say is on the technical implementation side itself, right? So a lot of times startups will, you know, hire, um, individual, you know, contributor or contractors. That are engineers because it is, or appears to be more cost effective. Right. Um, and then what happens is.

Andrew: There's not that person who can translate what the business outcome is to that team. And so there's a lot of decisions being made that again are going to just slightly miss the target. You stack up sort of those decisions over time and you miss the target by a large margin. Um, and then, you know, a big piece of this that I like to, Tell a lot of the founders that I work with is you don't have to solve every single problem today.

Andrew: You think of startups as [00:18:00] like, you know, being on a pond and you're trying to jump from lily pad to lily pad and not fall into the lake. Right. And so that's, you know, the next milestone that your business needs to reach to survive, to survive a little bit longer, get a little more product market fit, raise more capital, generate more revenue.

Andrew: Right. Whatever that those milestones look like. And so as a, as a, you know, as a seat in that the person in that CTO role, right, is their job is to translate what decisions can we make that we have to make today? And what trade offs can we make today to make to that, to make it to that next milestone. So our startup doesn't die.

Andrew: We don't run out of money, right? We don't, you know, we don't have to raise more funding and sell the company or sell parts of the company prematurely. Right. And so, right. Right. Those are little ways, it's a complex task, but those are little ways that you mitigate projects, uh, project failures along the way.

Andrew: And then I'll add to that a big piece of this, just like purely tactical, is if you're working, whether you're working with a consultancy like Vertice Labs, uh, or you're, you have an internal team that you trust, you know, right, [00:19:00] no matter what, You use your product, test your product, right? So part of our standard process is, um, we, you know, kick off a project with a client.

Andrew: Um, we go through the, you know, alignment exercise where we're, we want to figure out exactly what your goals are. So new product goals are pretty simple. It's like, get it built, get some people on it, get some feedback, you know, test the product market fit, and then grow it a little bit from there and test again, right?

Andrew: Um, And so, uh, I lost my train of thought.

Andrew: Um, I was saying, oh yeah, so the standard part of our process is, um, so we start with that, right? That, that alignment exercise. And then, um, really early on in the development phase, uh, of working with clients, we set up a test environment for them to log into. They touch it, feel it, get feed, feedback from stakeholders who are, you know, on their side.

Andrew: We'll give them, you know, constructive feedback on the direction of it. But the biggest piece of that is, um, what I like to say is, if your technical team says something is [00:20:00] done, but you can't use it, like you can't go through a full, you know, test of that, that feature, it's not done. And so you shouldn't consider it done when you're planning.

Mehmet: Right. Um, very, very insightful, Andrew, really, um, on, on the point of, you know, because you mentioned, You mentioned a couple of points, which, you know, like, uh, stuck in my head, as I, as I can say. Um, there is also like this, uh, fact, especially in startups that, because especially if they are maybe VC funded, so they have to you know, move fast sometimes, right?

Mehmet: Um, and moving fast means also like taking decisions. So, and, and here is the famous, I would say dilemma, like, should we all rely on data alone? Is it like mix of data and gut feeling? I like to call it this way. So, [00:21:00] You know, like how, how again to do this balance because sometimes, you know, like really you need to take decision and you need to move very fast.

Mehmet: And this is both, you know, that it might be a technical decision that affects the whole business. Uh, it might be, you know, adding a feature. It might be, I don't know, like, uh, um, changing the technology stack, whatever it is. So we need to, to, to take decisions fast, but at the same time, you need to make sure that it's like, It's, you know, taken based on some facts.

Mehmet: So what have you seen working in this space?

Andrew: Yes. Uh, this is one of my favorite, I actually just wrote a, uh, uh, an article on this. You can check it out at vertce labs. io. Um, but basically I titled it, you know, moving fast without breaking everything. Um, and, and, you know, this is one of the reasons that I love the startup space and the S and B spaces.

Andrew: Um, you have to move fast because, you know, every day that goes by, right. You burn more capital, you risk failure, you risk, you know, you know, first mover advantage, things like [00:22:00] that. Um, and so I, I like to think of it as like this, right? In the early stages, you have an idea and you have a pretty good reason why you think that idea would work, right?

Andrew: Um, ideally you're solving it as a very specific pain point for your target customer. Um, and so you have to build something to test that idea, right? Now you're going to start with, you know, interviews, customer interviews and getting feedback and testing the market. And you certainly should because that's.

Andrew: Generally free. Maybe you spend a few dollars on some gift cards, but you know, it's not going to cost you a thousand, hundreds of thousands of dollars. Right. Um, and so you get that feedback and then you go to, and you go, you get, you test what you think will solve that problem. So you build the first iteration of your product.

Andrew: The key to that is you don't overbuild it, right? You don't build too much at one time. You build enough to solve a very specific pain point, right? Then you release it into the wild. And that needs to be. It needs to solve the pain point. It can't be like a premature product that needs to be solved. But you also don't need to solve all the tangential pain points that come along with it.

Andrew: All those, those are, those are very clear [00:23:00] opportunities to scale, uh, your product in the future, right? And so that's one way you mitigate, you know, not having all of the data yet to make that decision. And then you take it that this process expands, you take the feedback from that first iteration, you gather the data, you're doing your customer interviews for people who are using your product, right?

Andrew: You're getting feedback from them. You take all of that data and you use its usage patterns of product, right? You take all that data and that helps you decide very clearly what to focus on next. Um, now you don't just make the decision purely on data. You also use, um, again, you, you know, the problem space and the problems that you're trying to solve.

Andrew: So you use additional hypothesis that you formulate to pair with that data to make the decision on the next direction for the company. And so I've seen that work very well. We've scale up, scale up, start up from the idea of the 28 million in revenue that first year using this process. And that was a very fast, Break some things process that we use, but we hit the milestones we hit.

Andrew: We were [00:24:00] able to land on the next little lily pad every single time over the course of, I think, you know, it was over the course of 18 months. And so that's an effective way to approach that.

Mehmet: Wow. Fascinating, Andrew. Um, we can't be talking about all this stuff without talking about AI, right? So, um, so let me start with the AI.

Mehmet: Recently we are starting to see more and more, uh, AI being used for within the software engineering process itself, I mean like generating code. And so, you know, Is this like something which is now because it's been, I mean, for a while, maybe less than a year, I mean, in the sense of, you know, going mainstream, but is it like now that default de facto, I would say, um, way when, when, uh, even hiring maybe [00:25:00] the, the, the talents that we need to let them know that they're going to be ready to embrace, for example, working with AI technologies that can help them in, in.

Mehmet: Writing code. Is this now what's, you know, hot and trending?

Andrew: It is hot. It is trending, but it's not a trend. I think it's here to stay. Um, so I use AI tools. I love it. I still write code. I love to write code. Um, and I use AI tools regularly to write code. Um, I use, you know, I started out using GitHub's Copilot as an extension into the Visual Studio Code IDE.

Andrew: I now use Cursor, which is an IDE that has it built in and connects directly up to the LLMs. Um, and so if you're, and then there's, you know, Purcell has its, you know, has a generator, there's a, there's a ton of tools out there that just significantly reduces the mental load on any given engineer and on any given day, which is nice because I know what to build this [00:26:00] thing and achieve this outcome with the thing that I'm building, uh, and I know how to do it by hand and I could do it that way, but it's nice to offload some of that Um, the mental load of how do I structure the code?

Andrew: How do I write these lines to to a tool? And I can focus more on the business context and the problem that I'm trying to solve, which again, with the process of aligning vision with tactical, um, gets you to the solution, the outcome a lot faster. Now, there's a couple things to know. I would say, um, what I'm seeing is, uh, AI is not replacing people.

Andrew: Yet. I would say yet. I think, you know, maybe at some point it could replace, you know, uh, you know, engineers, but it won't replace them all. I don't think, you know, us as software engineers have to fear our jobs, um, going away, but they are certainly making us a lot more efficient. And so as an example, junior engineers who are like earlier in their career and still learning, you know, more complex software engineering concepts can produce a lot more, uh, yeah.

Andrew: Code [00:27:00] integrity is a lot higher. They can produce because they have a guide, they learn faster, and they can produce at a higher rate than previously before without using the tools. There is something to be said about the human in the loop. There are always, I think, always going to need to be humans, uh, involved in AI, at least for the next five to ten years, um, because someone has to provide the context, right?

Andrew: So the AI as it is today is not necessarily It's not, it doesn't have drive and that's a innately human, um, a human thing. Right. And so someone has to provide that context of why we care about the desired outcome and why we're, why we're doing what we're doing and what we should consider, um, as a part of what we're doing.

Andrew: So that's a big part of that too. One of the things we've started playing with is, um, automated, uh, AI agentic workflows, and so that is, um, AI agents, you know, like every role on the software engineering team, you can build. You can train an AI model to do right. So, [00:28:00] you know, you have QA engineers who are writing tests and manually testing the product.

Andrew: You have DevOps engineers who are running infrastructure as code and, Scaffolding your, your infrastructure in the cloud, you know, back end, front end software, software engineers. So every one of these roles can, um, a model can be trained to play, uh, that role and interact with each other. And so we have started using AI agents in our SDLC process to see, one, how fast and how capable these models are.

Andrew: Uh, at doing these types of activities and then the quality of what the code is that they produce and what we have found is, um, you still have to, again, you still have to have that human in the loop to provide the context. You sometimes you have to settle tie breakers between AI agents and they kind of get in the lock.

Andrew: But for the most part, they can produce really high quality code at a significantly higher velocity right because it's an AI model, right, it's using compute resources and not mental resources and they're not [00:29:00] they're not hindered by Um, the scale of linear time, which is a whole nother argument that I can get into whether time is linear or not.

Andrew: Uh, so anyways, uh, I think it's quite a powerful technology AI as a whole. Um, I think humans are still highly required to be involved.

Mehmet: You know, like, I'm happy you brought the, you know, the topic of agents because this is This is what I'm personally excited about since more than one and a half year, I would say, when the first, I saw the first attempts, uh, with something like, uh, auto GPT and, uh, baby AGI and, you know, these, these other projects, which are in my opinion, uh, it, Maybe we're not there yet to your point, a hundred percent.

Mehmet: It opened our eyes and you know, the community eyes that this is something which is the future. And I was so happy when I saw also like, uh, I think it was last week or the week [00:30:00] before, when I see Anthropic, when they released the new cloud model, they, they introduced also, they start to talk about it. My, my expectation is.

Mehmet: Open AI has something that is hiding from us and waiting the right moment to announce it. It's just a feeling that I have because they couldn't have figured this out. I don't believe that, but yeah, it's a game changer. And I think, but again, like to add to your point that I like this because on the creativity part, this is where, you know, Someone like you, Andrew, will always be needed.

Mehmet: A person who has this, you know, experience and, you know, you can Not only, I mean, write code or advise them like who the talent that you need to do, like you become kind of the orchestrator for all these innovations. And now instead of thinking, [00:31:00] for me at least, instead of you, As the expert, Andrew. Oh, should we use this?

Mehmet: Should use that? Like the code is not optimized here, so you just focus on, you know, the innovation and the creative part, which it's I'm excited about it, honestly. And I think, you know, this is this is what would allow. Which is good news for you, by the way, it allows for more people to get more startups.

Mehmet: Like I believe we're going to have like more startups down the road, which is amazing now.

Andrew: Yeah.

Mehmet: Yeah. It's really exciting. Yeah, I do. It's

Andrew: really exciting because it does, it reduces the barrier, the cost of, you know, the barrier to entry. To launching an idea, right? If you have like a, a sound idea that you think solves a real problem, it reduces the barrier to entry, you know, primarily from a capital factor to launching a product into the market with, you know, you know, the engineers becoming more efficient, being able to supplement the process with, from an automated fashion, um, and, and then bring people up to focus primarily on the product space itself, [00:32:00] right?

Andrew: Like significantly reduces the barrier to entry, which is good for me because I, I love, you know, one of my big P's is I don't have I can build anything. I can't build a rocket ship, but I can build any piece of software. I don't like building software that, that people don't use, right? Cause I've built it for so long and I've worked in, you know, large companies and small companies.

Andrew: I like solving problems, right? And so that's the thing that I'm passionate about, which is part of why I started Vertiz Labs was, you know, I want to solve real problems that make an impact on people's lives. And so you get to focus on that, which is really exciting. So yeah, and like, I, you know, they, I think, you know, Anthropic released a new version of Cloud that now has the capability to interact with a computer the same way a human would, and it can drive around, browse the web, it can solve, you can give it a, you can give it a problem that is a multi step, multi task problem.

Andrew: And it can go to websites and figure out what it needs to do on that website to gain information and submit forms And like it can use it's starting to use computers like humans do and [00:33:00] that's that's really interesting

Mehmet: It's really cool. So I didn't try it yet. I saw just the video and it's mind blowing, right?

Mehmet: So yeah endless endless opportunities, I would say now from Again, AI perspective, but this time not within the coding process, or I mean, software engineering perspective. And maybe this applies not to start, it might apply to startup as well. So, when and when not to use AI, Andrew? I mean, because we don't want to just complicate our lives and just say, Hey, I gotta use AI for the sake, like I say, I have an AI enabled, uh, Whatever, startup or like we leverage AI to do one, two, three.

Mehmet: So what, what really matters here?

Andrew: Yes. So that's a really good question. It's something I'm, I've been diving into deeply lately, actually. So, um, so AI is thinking of it like this, it's a, it's an ingredient, right? [00:34:00] So like, you know, if your product is a restaurant and you serve, you know, food, you serve dishes to your patrons.

Andrew: Um, AI is an ingredient in one of those dishes. So it's like, you know, salt or parsley or garlic, right? So it's not, it's not a product and it's not a solution. Right. You use it as a part of your process to enable your engineering teams to move efficiently, to increase the customer, you know, the customer experience, the customer value of your product to better deliver on your services.

Andrew: Right. And so you still have to solve a problem. Your business still has to solve a problem that people will find value in. And then AI is a tool in the solution of your problem. Um, it is incredibly powerful. And it can do a lot, but you, you can't start with AI as being the solution, right? Because it needs to solve that problem.

Andrew: So the way that I, I'm actually, I'm actually about to launch a AI workshop to help with help businesses figure out the best way to start with [00:35:00] AI in their business. And one of the key aspects of that is Take a problem that you're already solving today, right? You're solving it maybe with humans, maybe with some, you know, some code that does some automation and take that problem and then ask, how can I solve this better, faster, more efficiently?

Andrew: Uh, and then ultimately what the impact is to the customer, right? And if you, if you look at it through that lens, then, then you have a real use case for it, right? And then you can start seeing the gains, right? The benefits of the company. And that is, you know, Increase revenue, you know, decrease, uh, operational costs, uh, and, and more satisfied customers, right?

Andrew: And so you can draw a direct line between those things when you start with the problem itself and apply AI to it. So, um, I've seen it used a number of ways. So products, you know, we'll use AI for, you know, you know, data transformation, right? Like if you have manual processes where People are, you know, manually entering data or crunching data.

Andrew: Um, and, you know, they get it 80 percent right. You can apply AI [00:36:00] to that and it will significantly speed it up. You don't have to hire as many people as you scale um, to do that transformation. Um, and there's process automation so that you transfer the data, then you can automate and enter in the process that you apply.

Andrew: AI, AI at different steps, right? Um, and then, you know, we just talked about the agentic world where that's kind of, that's coming up now. I think Not too far away from that, where you can deploy agents to do that on your behalf automatically. So, um, and then, and then embedded in products, people are using it for, uh, increasing customer satisfaction.

Andrew: So like, you know, you, the way that we've used it is there's the common, which this came up. That was exciting a year ago, but RAG applications where you build an AI workflow, you ingest some data, you synthesize that data, you can explore that data through chatbots, which was sort of like the, the, the entry into the world of what AI is capable of, but there's a bunch of other things that you can use that data for.

Andrew: So for example, you can use it to, um, especially with the public [00:37:00] LLMs, to learn more and add data into your product. That AI is already they already know about because they've been trained on such massive volumes of data, right? So you can add that data to your product where you otherwise would have to go Connect with apis and build partnerships with other companies Um that you don't necessarily need to do anymore because the AI knows it you can just ask it and this is part of your 20 bucks or Uh, or, you know, paper usage, uh, plan, right?

Andrew: And so, um, you can use it for, uh, understanding images. And so, one of my clients is using AI, uh, where, uh, the product allows you to take photos of things. It will tell you what that photo is. It will give you information about that photo. Uh, the information that you're particularly interested in and then give you, um, actionable insights.

Andrew: So, things that you need to do or can do to, uh, to accomplish the goal you want to do in their product.

Mehmet: Right. Uh, Andrew, like, you know, I think, you know, the discussion can, can go and go and go and go, but, uh, you know, we, we, we have to, uh, uh, to, to, [00:38:00] to, you know, like, uh, get it to, to hear now. I really enjoyed it.

Mehmet: Like really, it's, it's something, what, uh, what I was looking for from long time. I see the energy and this is why I want to ask you, Andrew, how people can get in touch with you? Where can they find more about you and about your company?

Andrew: Oh, yes. Great. It's uh, verticelabs. io. It's spelled V as in Victor. E R T I C E labs.

Andrew: io. Or you can email me at duncan at verticelabs. io.

Mehmet: Great. So for the audience, you don't have to follow letter by letter. I made your life easy. So if you're listening on your favorite podcasting app, you will find the links. Uh, in the show notes, or if you're watching on this on YouTube, you'll find it in description.

Mehmet: Uh, Andrew, again, thank you very, very much for being with me here today. I really appreciate, you know, the insights, the experience that you have shared with us, and also like the, how you gave us, you know, these real examples from real [00:39:00] scenarios in, in the real world as well. So thank you for doing that. Uh, and as always, this is what I tell the audience at the end.

Mehmet: If you just, Discovered this podcast by luck. Thank you for passing by. If you enjoyed, give us a thumb up, share it with your friends and colleagues and you know, share with anyone who you know. And if you are one of the people who keeps coming back and you know, send me their suggestions, comments and feedback.

Mehmet: Thank you for doing so. I really read and enjoy all of them. If I have to change something, whether good or bad, tell me about it. I will try to do it. And as I say, as usual, We will meet again very soon. Thank you. Bye. Bye