Struggling to scale your tech team? Not sure when to move beyond an agency and build in-house? In this episode, we dive deep into engineering team optimization, startup scaling challenges, and the future of AI in development with Vygandas Pliasas—a Fractional CTO and expert in engineering management.
🔍 What You’ll Learn in This Episode:
✅ When startups should transition from an agency to an in-house team
✅ The common traps founders fall into when working with external dev teams
✅ The step-by-step process for fixing tech bottlenecks and improving team efficiency
✅ How to scale remote engineering teams without chaos
✅ AI’s impact on development and how to integrate it into your tech strategy
✅ The biggest mistakes startups make when hiring developers
✅ The red flags that indicate it’s time for a change in your tech team
🎯 Key Takeaways:
📌 Miscommunication with agencies can slow progress—here’s how to fix it
📌 The right moment to build an in-house team and hire a Fractional CTO
📌 Scaling tech teams requires balancing technical leadership & business goals
📌 AI helps but doesn’t replace human decision-making—how to use it effectively
About Vygandas:
Vygandad is a seasoned Fractional CTO and software development executive with over 18 years of experience. He specializes in scaling startups, building dedicated remote teams, and driving technical innovation. With expertise in SaaS, e-commerce, and compliance-driven industries, he helps businesses align their technology with strategic goals, ensuring impactful, sustainable growth. From mentoring engineering leaders to streamlining operations, he empowers teams to deliver excellence while cutting costs and boosting performance.
https://www.linkedin.com/in/vygandas/
⏱️ Episode Highlights & Timestamps:
00:00 - Welcome & intro to Vygandas Pliasas
02:30 - Why founders struggle with agency-based development
05:15 - How to know when to bring development in-house
08:40 - The challenges of scaling remote engineering teams
12:10 - Hiring & firing: How fast is too fast, how late is too late?
18:45 - AI in development: What works, what doesn’t
30:10 - The biggest startup tech mistakes and how to avoid them
38:45 - Final thoughts & where to find Vygandas online
[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 Vygandas. Vygandas, thank you very much for being here with me today. The way I love to do it is I keep it to my guests to introduce themselves. So tell us a little bit more [00:01:00] about your background, your journey, and what you're currently up to, and then we can start the discussion from there.
Mehmet: So the floor is yours.
Vygandas: Hello. Thank you. Uh, it's, uh, it's very cool to be a guest in your podcast. My name is Vygandas Pliasas uh, I am, uh, say fractional CTO and ma management consultant. Uh, I have a wealth of experience of about 18 a over and 18 years, uh, and been working with the United States, United Kingdom, uh, startups mostly, uh, and.
Vygandas: What I do is I help founders and execs to streamline the process of product development, engineering, engineering management. So basically try to disrupt a little bit old processes and look from the different angle what's going on in the company and like in the whole organization maybe. and find what can be optimized, uh, [00:02:00] how to maybe, you know, sometimes reshape the departments and make everybody the most efficient, uh, you know, the most efficient that they can be.
Mehmet: That's, uh, great, uh, Vygandas and thank you again for being with me. It looks like, you know, we're going to have a, uh, exciting discussions because I think, you know, this is, uh, one of the main points. We try to cover, you know, on the podcast about, you know, everything, especially I believe you work more with a non technical founders, right?
Mehmet: So this is where maybe it will make sense. Yes. Right. Well,
Vygandas: technical founder. So it's like non technical founder or they don't have time. And they're like go figure it out
Mehmet: That's that's you know, very very good use case I would say and uh Now maybe a little bit maybe traditional questions. [00:03:00] You must have seen a gap right in the market Uh, so you have decided this so let's Take it in a different way Let's talk about the challenges and you know, the, the traps, let's call them, that usually, uh, such founders, they, they pull in so they would need a hand to help them.
Mehmet: So let's start from, from the main challenges you see usually when they come to you.
Vygandas: The usual, uh, case is the company, uh, hired an agency and then let's say they have a call once a week, something like that. The agency has a project manager and the team. So the communication is usually with a project manager and CEO or founder.
Vygandas: Tells all sorts of things that they need and they work on [00:04:00] it. So it's basically, you know, hourly based effort, you know, for the whole team and whatever they say is priority. Those people try to do, uh, so that's basically the usually like usual case, how it gets stuck. Then the problem area is that, uh, miscommunication.
Vygandas: Founders, like whoever basically interacts with the agency tends to have a lot of ideas and a lot of priorities that change really frequently. And, you know, like those agencies tries to optimize the profit margins also by throwing in more junior developers rather than seniors. So. There are a lot of misunderstandings, a lot of incorrectly implemented features, and it costs a lot, and the founder, or whoever [00:05:00] is interacting with them, starts to become frustrated, irritated, and, like, unhappy, because, you know, stuff does not get done.
Vygandas: Done simply. Uh, so that's a usual case I'd say. And that's another case when they just need a view from different angle and somebody with more wide experience to come into the organization, talk to head of product, head of engineering, head of. operations sales support and so on so on and basically figure out what is the current situation find where the vision should lead and then make a proposal on what could be, uh, changed.
Vygandas: So basically prepare the plan for somebody to execute the change management transformation within the company.
Mehmet: [00:06:00] Cool. Is there any specific, uh, phase where you see this happening more? Like, is it like after maybe they created an MVP and they started to go out and then they want to really actually build, uh, you know, more features around, you know, the MVP or do you see it?
Mehmet: No. Starting even from the MVP phase where they, they have these struggles.
Vygandas: Well, so it makes sense to do the MVP with an agency because it's easier to estimate, especially initially. So you have a spec and then they tell you the price and you agree and you get it. But what happens later is that they switch to hourly because they need a lot of improvements, changes.
Vygandas: So it doesn't make sense to write statements of work or, you know, and just throw [00:07:00] in requirements like Jira tickets or Trello or something like that, or just emails to the PM. And. Then, and then it gets more complicated and there appears to be more complex, um, features. And that complexity becomes a bit too hard to handle for the agency people because Many, like, you know, in my case, with those whom I work, those agencies, many tends to give the devs multiple projects.
Vygandas: So they work on this client, on another client, maybe something else, maybe fixing some legacy stuff. So they don't really They are not 100 percent into that single project, as if we're like full time contractor or employee would be. So, you know, that starts to cause all sorts of problems. Let's say a PM leaves.[00:08:00]
Vygandas: So, nobody has any sort of idea what, what, what is this project about. Maybe some devs don't really Like kid that much or had too much work to do and they just are taking the ticket getting it done next taking ticket maybe some different project and and like the Shuffle all of it and then new pm comes in and like all all the cycle starts again And so, in my case, uh, also what I noticed is that usually it's not, um, like originally a SaaS company, but rather than a company that already had some sort of business, maybe even like old fashioned business, and then they created some software for themselves.
Vygandas: And it turned out that other people, other companies make competitors. Also need the same solution and they think, okay, [00:09:00] let's convert into that, uh, SAS model and allow everybody else to pay us and like use it. So, you know, otherwise they will make it themselves anyway. So better to take the revenue that way and they have the budget, but they need efficiency and speed.
Vygandas: Right.
Mehmet: Now you mentioned about agencies and of course the, the challenges that is caused by the, not because of the agencies are bad. Don't get me wrong guys, if you are an agency owner, but I mean, Yeah, you mentioned something, and this is understandable from the business model of an agency. Now, the question is, when is the right moment, because you mentioned about, okay, MVP and later, but, and this is, I think, something you help with, Vygandas.
Mehmet: I need to take now, let's say I'm, I'm, I'm the founder, I'm the CEO, and now I need to take the decision to hire these people [00:10:00] as, you know, part of my company. Um, are there like some minimum, I would say signals or requirements that should tick the box before I go and make this move to bring these guys in house and get someone like yourself also on a fractional level to manage them or, you know, it's, it depends like what are like the different scenarios and options that I might have as a founder, uh, who having these challenges you just mentioned to make sure that, you know, I fix things.
Mehmet: Right. So what's my journey looks like?
Vygandas: Okay, so definitely agencies are not bad. Uh, I mean, like, you should start considering the switch when you get frustrated, when you don't get the speed that you would like, and when you You basically you start building the product. So you want in house [00:11:00] knowledge. You want to start writing documentation.
Vygandas: You want people who would care only about that product, like, you know, give t shirts and stuff like that. And. Another component, which is really important to this equation is that you want to optimize on the budget, because like if you have like a lot of cash to spend, it definitely can go with an agency.
Vygandas: It's fine. You will get what you need. But many times it is significantly better like deal and the better value to start building your account. Own teams that would go under your own flag, right?
Mehmet: Yeah. Yeah. You want to say something? I
Vygandas: think getting frustrated, need more speed and won't optimize on, on spendings.
Mehmet: Okay, cool. So, so, so you come when [00:12:00] actually, if I'm the CEO, I'm feeling, you know, I don't know what to do next. Right. So. Uh, if I want again, because what I'm trying to do with you again, this is I want to create kind of your ultimate customer or your ICP journey, which is, you know, an overwhelmed, a frustrated CEO.
Mehmet: So how the plan works like, um. Of course, I know it might depend on the situation, but there must be kind of a maybe playbook or like, is it like, uh, you start certain assessment first? Like, is it a kind of some questionnaire you hand over to the CEO, how, how usually this works? And, you know, of course I believe like later you come with some actionable plans.
Mehmet: So how, how exactly the, the, let's call it the process works. Sure. Um,
Vygandas: so there is no. At least in my playbook, no [00:13:00] like strict set in stone questions, but there is a framework how I approach it. So I start talking with the CEO, so basically, what's up, what's the problem, what you don't like, what you want. And then, uh, I talk to all the other people like, like level by level going deeper with the most relevant people, of course.
Vygandas: And then the, let's say in the chain of command, the further you go, uh, you get completely different perspective. The higher it's more. Politically careful, like, politically, politically correct. It's like, vague words, you know, nobody wants to hurt themselves, soccer league. They know how to, like, tell nothing with a lot of words.
Vygandas: Stuff like that. And then you dig deeper, deeper, deeper, and then, you know, you come to the, let's say, lead developer [00:14:00] or product manager, and they tell a completely different story or something, you know. The requirements are not clear. There's like too many of things changing. There's no No, no process, how to approach things, uh, all sorts of things.
Vygandas: Maybe people are leaving too frequently and so on and so on. So all sorts of things, maybe it's a small startup or maybe it's just a, it's a big company with us. Like rising project on the side, so it depends like on the size, but you get to get down and then you start going up, you know, and you join to the call and say like, all right, so I've been hearing a few things about, you know, uh, that there might not be like clear process on how all those ideas.
Vygandas: Like gets born and gets transferred into development and like, you know, getting [00:15:00] shaped and what needs to be done and then taken. And that might be causing frustration. So it's like true or false. And, uh, you know. Get the answers, get the perspective, going, uh, step up, up, and then talking to those guys, and then, uh, talking to the CEO again.
Vygandas: And, you know, so you get the different picture, and then I write, like, a report. Uh, what I found, you know, what's the general feedback. Uh, how it's all understood within the company, within different departments, you know, like engineering product, uh, maybe sales operations. And then, like, I, I start writing some proposals.
Vygandas: I know that those proposals might not be right initially. And then I give the document to all the people I've talked to and they read it and then I go again and I get better answers because for those, uh, where the problem wasn't [00:16:00] the one that I described in the document, they will start saying, Oh man, it's like, don't do that.
Vygandas: It's, it's the only thing that works well, you know, something like that, basically uncover. It's like. Iterations by iteration by iteration and you uncover all those secrets, you know, lying deep in the organization.
Mehmet: Right. Now, so let's now shift to the other phase. So now, you know, I figured out I cannot do it alone.
Mehmet: I need some help. So you came in, you joined me. I'm just making this up. Uh, now I believe the next step, of course, you know, putting these, uh, Documents and, you know, talking to the stakeholders, but let's get into the technical aspect, which is actually with the people who are going to build the product now from your experience, and you have done this many times, you know, and your profile talks a lot about this.
Mehmet: [00:17:00] Now, how is your approach, you know, shapes? When going from a small team and now we're going to scale. So you came, you cleaned the mess. Now we need to take it to the next step. So here, tell me a little bit more about, you know, your approach in scaling engineering teams and making sure that everything works perfectly.
Mehmet: I understand sometimes even for some companies, they would have very dispersed teams also as well. They might be in different continents. So, Let's talk about this aspect here.
Vygandas: Well, most of my experience is built on remote teams and remote teams management and building. Uh, so that's the easy part, I guess, for me.
Vygandas: Um, it's tricky. It is not. So you need to approach it with care and think what you say [00:18:00] and on. Don't rush with decisions. I need to talk with everybody and find out what they think, because that's like a very strong human component, like human emotions and ego component there. So there might be somebody super strong in.
Vygandas: technology, like leading, like the development and everything, and there might be two ways. The person might be super good architect, or they might be really good manager. Like, if you make a super good architect into a manager, they will be overwhelmed because they, they will try to do perfect job in technology, but they will be buried in like managerial tasks, basically.
Vygandas: Talking with people and not doing anything hands on. Or it might be, like, vice versa, where the potentially [00:19:00] great manager, who gets along really well with everybody, would be really desperate if they were left in as a Dell, for example. So need to find out the potential and even like, let's see, somebody sees somebody potentially becoming like a engineering manager, director of engineering.
Vygandas: It's a bigger team. You need to make sure that they. want that, really, and they understand how it will change their day to day. Because, like, if you get 40 people, like, there's no way you will be coding. Like, you can, but it will be, uh, like, it will be some sad time, I think, if you want to, if you want to get all the things straight.
Vygandas: Um, and then, you know, we start [00:20:00] proposing ideas, like carefully, like, what do you think? Who would be the good, like, person to handle all those, like, nitty gritty things, like communication with, let's say, product, like, with other departments and basically shaping it so you guys could, you know, work with it.
Vygandas: Like, not saying, like, who could be your boss? Like, that sounds bad. Approaching it with like a smart angle and getting some ideas and after also again multiple iterations. Everybody is fine with it. Like either you need to hire somebody new, like a director or engineer manager or leads or like extra devs.
Vygandas: And there's somebody who could handle those teams, but need to make sure that it's the right person who will be, let's say, team lead also, otherwise, you know, you will give them for, for other new developers [00:21:00] and they will be burned out like in a couple of months.
Mehmet: Right. A quick question. Uh, we got this on.
Mehmet: You know, hiring and firing here. So, and we hear a lot of, uh, opinions. So how fast is too fast and how, you know, late is too late. Let me ask it this way.
Vygandas: Um, I'd say you get a red flag, you mark it. And watch it, uh, no rush decisions, uh, you get strike two, you talk, you try to figure out what's wrong if, if it's like strike three, like out, you know, you can put a person on pip, but usually like, uh, it ends up the only one way. So [00:22:00] it's like, don't wait for too long because if there is like repetition or the pattern of behavior, No point, you practically on strike two, you can start practically put a job description already.
Vygandas: Um, so it's, it's difficult. It's, it's never easy, but if you wait for really long and like hope. For something to happen. It will not happen. Like you have to do something about it. Of course you need to talk. It might be something silly. You know, it might be something small, something silly. So you can't figure it out.
Vygandas: But many people are afraid to do that because it's like, sounds bad.
Mehmet: Right.
Vygandas: Also, a funny thing is, uh, high level managers usually knows one thing if they see a consultant. [00:23:00] Yes. That it happens that execs hire consultant so they could blame the consultant for the layoffs It's never good a sign, you know to see a
Mehmet: consultant I think Absolutely, yeah, yeah, I heard that story similar to this now Talking about you know, and of course cto show.
Mehmet: So managing to Keep track of of course The tech stack, but at the same time keep making what the business requires, right? Um, so my question to you, uh, Vygandas is like how You know, is there any framework to make sure that? Everything that we are doing on the technical side of the house Stays within [00:24:00] the business goals And, you know, is there some signals that, you know, start to appear, maybe that flex, let's call them this way that, Hey, like you, you need to go to, to tell the, the business owner, the founder, CEO, whoever, like, Hey, I think there's some mismatch here.
Mehmet: And I think you need to rethink this again. Right. So what is. you know, the best approach, what kind of experience you can share with us regarding that?
Vygandas: Well, uh, what structure I usually implement is, um, you see product has to build their own roadmap and has to be Well thought through and prioritized. So the, how to make a list and then come to engineering and then engineering will tell what is hard, what needs to be split, potentially what could be [00:25:00] split.
Vygandas: Uh, so work with a product together and, and, and, and make those evaluations. It could be a low level of effort. Um, t shirt sizing. That's usually the easier one, like story points. It's crazy. It's because it's at high level, long initiatives, especially portfolio initiatives that takes quarters or half years and, and, and, uh, get it planned there and get engineering input all the time.
Vygandas: Also the same time engineering has to build the very same kind of the roadmap that needs to be done on. Engineering side, like technical, uh, do they need to, let's say, rewrite, uh, monolith structure into microservices? Do they need, uh, implement? Kafka, let's say, or auto scaling, they need to do some sort of migrations, lib changes, version updates, uh, other things.
Vygandas: Maybe it's, uh, Java 8, [00:26:00] for example, if they would like to upgrade a little bit. Oh, and, and, and then found the best. Way to have like 80, 20 is engineering and basically do all those improvements at the same time. Also work on, on, on the product, assuming there's like a larger team, you know, who can handle like, uh, then in an example, then those two making, uh, infra upgrades, eight making product stuff, let's say, and, and.
Vygandas: If you do it right, if you agree on product roadmap, you estimate it more or less well. And if you do the very same thing with technical roadmap, you can plan it in and get the consensus among all the leadership and can start working on it.
Mehmet: Right. Now a question, maybe not very [00:27:00] much as a follow up, but it's something that just popped up in my head.
Mehmet: So Um, I know that you work with different also kind of startup and also like, uh, uh, enterprises. How much, you know, the vertical they are in, uh, shape also your style and your frameworks or is it like just, you know, SAS similar to e commerce similar maybe to, I don't know, maybe FinTech. So how things might change when, when you work with different type of startups?
Vygandas: Well, uh, most experience that I have is in travel. Uh, well not like travel, like more like an e commerce style or getting the users into the project, but it's a lower risk where you have massive marketing budget, massive traffic coming in and you basically need to Figure out all, how, how to get it done fast and like, uh, react to market changes or [00:28:00] all of that stuff.
Vygandas: And the other kind of, uh, projects I've been working is like in healthcare, which is super dangerous. Somebody can die if you make a bug, you know, that kind of stuff. And there's a completely different approach. So in where you need to be super fast and react, it's better to Take, um, say make, you can make more mistakes.
Vygandas: It won't hurt you that bad. You can move faster. You can have shorter QA process. Uh, it really works if you hire really strong senior developers. You might not even need manual case, uh, testing what they've done. It might be CQA engineer who's like training everybody and making sure that they actually testing it themselves.
Vygandas: But on the medical stuff, it's like you have to test everything and you have to do regression tests before the [00:29:00] release again, which takes a lot of time and, you know, to be. Compliant with all the regulations. So if the case says, Oh, that smells basically dropping it off and like putting it back into and look into it, what's, what's wrong with it.
Vygandas: It might be not only like engineering stuff that was not incorrect, was incorrect. It might be from legal. Somebody felt it might not go right. It might be from product. It might be from medical, you know, it might just not. Be what you might think the Your as you stay holder wanted you to build so that many more risks like huge fines If you make some something dumb,
Mehmet: yeah, no make makes makes absolutely sense now.
Mehmet: We we can't Be, uh, sitting, recording an episode, especially with a CTO like yourself, Vygandas, without asking you about AI, [00:30:00] right? So let me ask you, how AI affected, you know, the way you engage with your clients?
Vygandas: It made it harder, actually, in some, in some It made it easier in another way. Let's say, so it's easier to, let's say, make emails, optimize it. It's easier to, let's say, document some things. Let's say you have notes. You can, uh, let's see, ask to IGPT, like edit, proofread whatever I had. It makes it a bit more readable if you wrote everything very fast to your iPad.
Vygandas: Uh, but what I found is like everybody is now bombarded on emails with an AI emails and requests and all of it's, everything's automated and it's a bit nightmare. Kind of crazy in LinkedIn, [00:31:00] like how many messages are coming in and it's all JHPT stuff. Um, then companies tend to be more careful, uh, with NAI because again, it's a third party.
Vygandas: So nobody really wants to give the the third party that easily. Uh, so also everybody hit the stories where airlines, uh, chatbot promised something and then they got sued for the chatbot's actions. Uh, so it's, it's more, more cautious, uh, on development side. I practically tell everybody to implement, uh, some sort of, um, pull request checking AI.
Vygandas: That's, that's, uh, that's an awesome thing. So you keep humans doing the pull request reviews, but you can also have an AI that can check your JIRA, that can check, uh, your code [00:32:00] and basically write the comments. What might be incorrect? What might be a better thing? So to basically. It makes the check like the very first iteration of the most obvious things, so that saves a lot of time for the human reviewer of the pull request.
Vygandas: Um, what else? It's easier to shape requirements, but it's like, it's not really that much. Of integrations yet.
Mehmet: What about coding? We see a lot of co pilots and I can name a bunch of these tools like Cursor, Replit, there are many others also. Again, is this something that from your perspective as someone in leadership makes the life easier or like harder for you?
Vygandas: I think [00:33:00] devs has to use it. It's like a tool You you you had a hammer manual now you have like electric one
Vygandas: Definitely but I always tell like You are still the head. AI is not the head. You're the boss, like the developers. You're the boss. You manage your tool. Like you cannot listen blindly what it tells you, think at every step, because it might make it even harder for you, especially on complex tasks. So what they found is extremely good when you need explanations.
Vygandas: So I experimented myself with a cloud. So it's like IntelliJ, uh, it works on all. WebStorm is JetBeans, uh, plugin for cloud Entropic and it can see your files and it can interpret really well the code base. I, I really like that [00:34:00] one. Of course, like there are a bunch of AI integrations like, I think, EasyCode, then Copilot, JetBrains has Copilot, GitHub has Copilot.
Vygandas: Uh, then the one, how it's called, um, I've heard that SiliconWallet is like 50 percent devs using it. Visual code. Yes. Cursor. Um, that one, but anyway, like that, I can give you an example where it will not work and it might mislead you and like burn a lot of hours is let's say you language probably doesn't matter.
Vygandas: Let's say Java, you know, uh, you need. To make a worker that would handle jobs, let's say it's like a reflection, you get a class, you get a payload, it like creates it and executes it, but then there's another constraint that your jobs has to be grouped, so one thread can take one group only. And execute that group at the same [00:35:00] time.
Vygandas: And then, you have automatically scaling workers, servers, based on workload. So, it has to be all synchronized. See, on Redis. So, good luck with that. Asking AI to get it done. It will take A lot of like AI loops, you know, it gets confused what you want because it's a lot of files already. It's like database connection, synchronization, thread synchronization, all sorts of things.
Vygandas: It seemed that humans still could have done it faster, but you can still ask AI to help on small things. So you need, it's not probably a good idea to ask it to do the whole thing, but it's better to ask it to like do small things. So you say, okay, I have these files, analyze it or those files and all the related files.
Vygandas: I need like this specific. Thing to happen that then it will be good. If you say like I'll just [00:36:00] do it all for me. It'll be We're
Mehmet: not there yet, I believe we're not there yet to to to get to this point and I agree with you because you mentioned Something you are the boss. You said you are the boss and I think yeah, we are all the boss not only for You know coding purposes.
Mehmet: I think Look, like AI is fascinating technology indeed, because it takes, you know, especially the repetitive tasks, you know, the, the thing that would take us long time to do and, you know, AI can do it in a much, much shorter time, but yeah, still, we need the human brain to be working and to giving, you know, I call it like, it's the, as they say in France, the chef d'orchestre, right?
Mehmet: So the, the, the chief of the orchestra needs still to be, to be there doing his magic, and I think this is. In this case, you'll be getting this. Right?
Vygandas: I'd say the developer himself, like Hersel, is the chief or [00:37:00] master. You can use, like, five AI agents to do, like, five things. But still, it's you who's accountable, responsible for all the outcome.
Vygandas: It's you who have the broad context, understanding. Of how it all needs to work eventually, in the long term, in the grand vision, you know, of what you're building. AI has no idea. It literally, like, does what you've asked for and doesn't really care about anything else. So, that's why you as the developer or manager, you are the head of it.
Vygandas: And you use it as the tool. And you, like, you wouldn't ask, let's say Like build me a house, like can nail a nail pretty well, but you know, it can't
Mehmet: actually do the whole thing. It has no idea what you want. Yeah. So I call it like systems 101. So systems 101, you know, if someone studied electrical engineering, they'll understand this.
Mehmet: So you have this box. It's a black box. So you have input and output. [00:38:00] So in order and output to come still, someone needs to feed the input. And this is where us, we are feeling the input. So a hundred percent, uh, we got this. Are we close to wrap up, uh, this episode? I still have to ask you two things. So first from a trans perspective, we talked about AI, but what I'm interested more, it's not like from.
Mehmet: technology. But what, where you are seeing like more, I would say, innovation happening in the startup space. Like, is it in FinTech? Is it in health tech? Is it still in the B2C space, which is like marketplaces and e commerce? Where are you seeing the most innovation still ongoing?
Vygandas: I
Mehmet: see a
Vygandas: lot in engineering related dev tools, because, you know, Devs, easy to build dev tools, so all sorts of helpers and tooling, uh, from my own like perspective, since I work more with healthcare stuff.
Vygandas: So I see really interesting healthcare implementations, [00:39:00] uh, helping, um, uh, clinicians to make decisions to not, to not missing something, uh, that you. With an AI, you could even make, like, pharmacodynamics checkup. Let's say you have a patient who was treated from, like, disease X, or they had a course of mix of medications.
Vygandas: But it's, like, a long term, right? A couple years, for example. And then you're prescribing a new meds, which, uh, let's say I knew. And for human, it's, like, it's very big context to grasp already, because it's a lot of Variables, but there you can feed all the pharmacodynamics, like pharmacological books into AI, like as a rag, like pre trained it already.
Vygandas: And then you can just say, okay, here's a patient, age, diagnosis, like [00:40:00] problems, meds, like frequency, uh, Adverse effects, whatever. And then we want to give this. And it can scan all of the, like, potential dangers and give you warnings. It should not make the decision. It's not a doctor. Doctor is doctor. They have to be responsible for that.
Vygandas: But it could flag, you know, potentially something that could go wrong, you know. So there's huge, huge potential for applications there. I even seen, like, uh, some other clinics for this. Uh, turns on recording when the patient comes in, like on the dashboard, and they discuss the problems, and it transcribes, and it sees the patient profile, it knows all the issues, and, you know, every doctor has to write down, basically, The note after the session with the patient and basically pre pre fills the whole thing for you and you just need [00:41:00] to add missing parts, you know, what didn't caught up well.
Vygandas: So it saves a lot of time,
Mehmet: right? Yeah. So again, fascinating stuff are happening and thank you for sharing your point of view. Now, last thing last, I would say, uh, Any final words you want to share and where people can get in touch and find more about you and your work?
Vygandas: All right, so final thoughts uh Embrace the ai keep learning.
Vygandas: It's it's going to happen. It's going to be the future Um, don't be afraid of it. Just learn how to use it properly Uh, if you want an advice help talk. Uh, you can find me on LinkedIn. You can find my website with the services that I do. It's core tip. com. On LinkedIn, it's Vygandas Blasus, like how you see it on, on the video, it's the same name.
Vygandas: Um, and [00:42:00] yeah, I was like, I hope it was informative and you learned something.
Mehmet: Indeed it was, Vygandas, and uh, for the audience, you don't have to go and search much. I'm gonna make your life easy because I'm gonna put all the links. Uh, to, uh, Uganda's website and his LinkedIn profile on the show notes. So if you're listening on your favorite podcasting application, you're gonna find them there.
Mehmet: If you're watching this on YouTube, you'll find them in description. Uganda, thank you again very much for being here with me today. And this is for the audience. Guys, if you just discovered our podcast by luck, thank you for passing by. I hope you enjoyed it. If you did, so give us aamp up. And share it with your friends and colleagues and if you are one of the people who keeps coming again and again Thank you for keep coming back and thank you for pushing the podcast to the top 200 list in multiple countries in the first Month of the year 2025.
Mehmet: So thank you very much for this and thank you also for for people who are in in dubai For making the CTO show as you know, [00:43:00] top 40, which is ranked as 14, uh, in the 40 most listen business podcast. So thank you very much for this. You know, it couldn't happen without everyone who support this show. So thank you very much for tuning in, and we'll meet again very soon.
Mehmet: Thank you. Bye-bye. Thank
Vygandas: you. Take care.