Looking at the future of business, Stepan explores how AI is transforming from conversational agents and back-office automation to hyper-personalised customer targeting. He unpacks the moral and strategic implications of AI, the skills needed for future workforces, and why leaders must act now to stay competitive.

Welcome back to The Growth Workshop Podcast for the continued discussion with Stepan Kopriva from Adastra. Stepan, where do you see you're getting greatest impact, and where do you drive the greatest impact for your clients in leveraging AI, the biggest impact that AI can have in 2025 in, let's say your biggest sector, the financial services sector?

Yeah, that's a really interesting question. So actually, in financial sector, I would, I do see several, several areas where AI can help a lot. The first one is for the businesses who are fully digital, I think the conversational part of AI, so the agents who are talking to the customer, being it just just as a text or being in the video agents, I think that is, that is a big one. The second one, I would say, is the automation in the field of software development, because the AI tools are very, very, let me say, experienced, or actually might be used in software development with the biggest impact 30% 30 to 40% of code in some organizations is already not in FinServ. However, it might be used in FinServ as well. Is already produced by AI. It depends what the teams are working on, actually, but still, it's a great one. So conversational software development. And then the third one, I would say, is definitely next best offer for the customers,

focusing on, on, on customer segmentation and actually predicting what, what one could, could could offer. And then I would add the fourth one, which is automation of the back office processes, if you wish. So any automation of processes, invoice processing, for example, that is a big one as well, and that is being done by AI on daily basis these days.

Yeah. Well, so what we're going to do now is we're going to put the question that Matt has asked you into an AI application. I won't name names, just because, you know, yeah, we don't want to be becoming sponsors here. So same question in there based around the financial service sector, as you mentioned. So interesting, I think very, very much aligned. So here are four bullet points that AI has given us, so Mr. Ai, to the actual technology, so we've got fraud detection and risk management would occur. So that's an interesting aspect around about how to reduce risk or detect fraud within financial services, operational efficiency. I think that's a pretty obvious one, but you talked a lot about operational efficiency. I think this is interesting. Again, similar to what you said, personalized customer experience, so maybe with a conversational intelligence aspect, as you referenced as well. And again, I think you said this, but this is just probably a little bit

more pointed. The last one is investment strategies and trading. So they might be able to do is more about the product refinement and actually positioning some of those insights. So Mr. Ai, is exactly on the ball.

I have another question about the workforce in a moment. But before we do, I think one point I think is really interesting. We think about the audience of the Growth Workshop Podcast. When we talk about growth, we were thinking, you know, traditional ways of growth, more sales people you bring into a business, the more out and more throughput you get. I think one thing you mentioned there, which is around segmenting and targeting your customer base, I think that's a huge opportunity in a whole manner of industries, B to C, B to B, in terms of using AI to unlock some more of that opportunity by being able to focus on where you focus. If you got any examples, obviously, I know you could, when we had to talk too much about live projects, but if you've got any examples of of how that works in practice.

Well, I think that if your data, or the data of the organization, is in place and in shape, which is very important for AI, because we are in AI, we are working with the with the data, and We are actually automatically. What AI is doing is automatically extracting and learning from the data. So data needs to be in shape. Then I believe, for example, in retail, in and in in electronical sales, so eshops and stuff like that, you can add huge value by learning from what the customers were buying. If you get, if you can connect the data, for example, with telco operator, which is not often the case, but I've seen that, then you can, then you can connect it with where they are moving and so on. And you can, you can segment them and then target each of the segments. And if you get it to the extreme, then the size of the segment is one. And you might have a strategy for for each single person who is buying with you. And it's doable today, because if, if you, if the

system is working automatically. Then it can create a strategy for each single person buying and provide individual offers, next best offers, maybe even, say maybe even, like, sales on some items and stuff like that. And it's very, very powerful, and it can be done given the parallelization of the of the computation of AI, and given the technology these days, it can be absolutely done on the on the individual level.

Can I ask a question? Because I think this is what's been going on in my own head, and I'm sure other people, because i i class myself as quite average. If I'm thinking this normally, I think others might be thinking it as well. I know, but I'm curious about what you've just said, because my opinion is that that's been doable that type of work for a number of years. It's just the scale in which it's been rolled out. But just before I go into the question, because I'd love you to answer that sort of I think that's been around. You could do that, but maybe there's a reason why it hasn't been done at scale. So if we talk about tourism, and we're in lovely Prague today, beautiful, you know location, walking around the streets before today, it was stunning. But where's your favorite city in the world, outside of the Czech Republic?

I will say London.

Oh stop it.

I studied my master's in London.

Oh, fantastic.

And I just fell in love with the city. So London.

So London's your place. What I'm assuming from what you've just shared there is that anyone that travels to London, in theory, should be able to receive here are your next three cities to go to based upon what you've just done, from spending stuff, what you've seen, and actually just the fact that London is a bit like Paris, but they speak very different languages, and the food's better in Paris, supposedly. But that might be the next recommendation from what you've just described. Is that correct?

Absolutely, absolutely.

I thought that's been around for the last five or so years, that type of work. Am I correct in thinking that?

It's been around. However it I would say that the devil is in detail, and that is in the granularity, right? So it's been around for clusters of people like you did the clustering of the customers, and you would say, okay, the size of the cluster is, let's say, 100 Why did you do the clustering? Because computationally, it was quite expensive to do it on the individual level, and as the technology is getting, let's say, more cost effective, then you can do this on the individual level. Also you need to collect the right data. So it's been around for some time already, but I think it's a it's a great example of where the technology did get these days.

That's beautiful, because that's the thing, because I'm seeing things going well, that's been around, but it's just because of the scale and the cost. So like medicine and the state having to provide medicine, or the corporate companies that own medical pharmaceutical firms, they're able to now, at scale, use AI to start to produce some types of medicine, because of the cost implication is lower, because you can do things much quicker?

Partially because of the of the of the scale and of the price, but also it's because of the new AI techniques. Let's say the generative AI, which is based on, or might be based on large language models, which and the generative word means that the AI can generate new results from the old ones, and they are using it in pharmaceutics, they are using that to create new sequences of of the structures. So, so, so in the past, the human had to say, Okay, we will go in this direction. We will try to generate new pill, for example, using this, this and that. However, having degenerative AI unlocks the possibility to generate millions and trillions of combinations automatically, without the human effort, and then test it possibly so. That's the generative power of AI, which has been around for five years, maybe even less, after covid It came in.

And I'm going to ask you a question about, probably a moral question later on, about AI, but in terms of the future of AI. And we know that people who have been on the podcast in the past, even actually, if you ask people, What do you think is going to happen in the future, humans find it hard to think about the future. We just have that disposition of staying in the current state or or before. But let's just roll forward another five years. What? What is, what's going to be happening in the world? Do you think?

Well, it's a tough question, given the time horizon, I believe that AI will actually, it will automate a lot of repetitive jobs number one. So the people who are doing repetitive jobs, and that doesn't mean just blue color jobs, it means white color jobs as well. So the people who are doing the repetitive jobs should be aware of that. However it's going to unlock, I believe, from. Actually the whole new space of jobs, which we don't see yet as of now, which where the people would be able to do completely new job based on AI, or to support AI and so on. And I think AI is just a technology. It's, and I believe it's, it's like invention of transistor, for example. So when people invented the transistor, it was a, in electronics, it was a, it was a huge, huge shift of the paradigm. In the very beginning, there were people who, for example, were using the transistors and so on. There were people who were repairing televisions. Now, nowadays we don't

see those jobs anymore. And in the very beginning, there were the jobs of people who were designing the systems using individual transistors. Those jobs are gone as well, because the industry just improved significantly. So people are now building the hardware using chips and so on, and those guys had to actually transform to learn something new. And the same holds with, for example, agriculture, when we had the automation getting into agriculture machines, way less people were required to to work, actually on the fields and and they could do something else, right? So, and I think with AI, it's the very same, just the magnitude of the change might be way faster. So people will need to, we will need to train the people to do some new kind of jobs. And probably we don't know, and many of those jobs as of today.

Yeah, and I read a brilliant article last week about what are the three skills or behaviors that people need now? It's cognitive agility, it's adaptability and resilience, because the pace of change over the coming months, not even years, is going to mean that we're going to need to utilize some of those skills, I'm sure.

Absolutely, absolutely. And you know, the other thing is that AI, actually, you don't need that many resources with AI to build a business, right? We've got these solo printers who are using AI, and they are making millions of dollars just by themselves. And there is something which didn't hold like 10 years ago, even in the software industry, where you were very efficient, you needed a team, probably and so on these days, if you are smart enough and you can do that, you might become a solopreneur, having your own business, doing millions of bucks per year just by yourself.

One day.

One day.

Just to round off this part. And Matt, I know you've got some questions around workforce and how it's going to impact that, but morally, if we take a step back, and Matt, I'd like your answer, and I'll go think of my own answer. But if we take a step back from business just for a second, and we think about AI, is still the topic, if you could use AI to solve a thing in the world. What would that be? And then this is prompted because the pharmaceutical conversation really so, you know, maybe not in a business context, but there's something out there that, you know, AI could actually solve for. What would that be?

It's a tough one as well. I would love to see AI solving the social problems and the problems in which actually partially were caused by AI, unfortunately, so So problems like the issues of social networks, where people can easily be influenced by some other people. In the old days, the sharing of information was guarded by the official media, like newspapers, television and so on. Whereas today, anyone can just join the social network and be influenced by small, a relatively small group of people, anyone can start a website and so on, and then do the advertisement on the social network, and that is super, super dangerous, I think, for the society, just in Czech Republic, because of how easily people can be influenced number one and number two, because of the amount of time children, for example, are spending on the social networks. There are the pieces of research which show how destructive it's for their psychology and so on. So if I

could, I would use AI to solve these two problems.

Yeah, I think such an important problem at the moment. And I've got a couple of young children myself, and they're not at the age where they've got smartphones yet, but I'm terrified. I'm terrified of what that means, because of the sort of the message and everything else and the and the way, yeah, the way people can be influenced differently. I also see amongst parents, and those in my parents, my parents age, age group, which is that, let's call them retirees, actually becoming as addicted as some of the young people. People, and again, you think about that messaging. And so I think, I think so. I think socially it's not an easy challenge to solve for but like, I think socially it's really important, and if AI can help that, I think that'd be fantastic. My answer is less people and more environment. I think if we can turn some of the power of AI into and I know there are organizations out there doing that today and using AI to to help you understand where you

can, you know, be more environmentally friendly in certain areas, whether it be the way that you hit your hat, you hit your home, or the way that you transportation, or whatever that might be. I think if we can start to leverage, AI, for the benefit of the of the environment that we live in. I think that would be, that'd be a really good

Yeah, I had a couple of ideas floating around that I'm gonna give two. So the first is that we work a lot in financial services, Matt and I, and the team and a lot of the organizations we work with serve the top 1% of earners. And I just know from personal experience how finance is quite a stressful environment. And if you look at the world and in the country that we live in, separately, is it's there's a socio economic difference disparity, right? How can AI, don't know, but support that to serve the wider spread, to actually, hopefully push that gap, and would be the first one, and the Secondly, I've pinched another one here, is being a dad. There isn't enough known quite rightly about children, but it's important to reference. I wonder to what extent an AI could support both the children and the parents in the future to actually make parenting probably even better than it already is. So yeah, that would be pretty, pretty cool to see how it helped. Don't

know? There we go. Thanks for answering the moral question.

Yeah, you're welcome

And sticking on this point of morals. And also just the future we touched on this just slightly in terms of the workforce and what that looks like. There are things you talked about Stepan, in terms of the types of roles and the agility that's going to be required, and that resilience, Jonny, to your point, but as a leader in a consultancy business that is at the bleeding edge, to use a engineering term, software engineering term, at the bleeding edge of of AI, what do you think, if you had to say there were three key changes in industry, and we're talking about the workforce, the shape of the workforce here. And let's again ask that question of fun of the finance sector, what do you see those being so this is on the people in that in that sector?

Well, I would say the first one definitely automation of back office. So, so that is, that is the huge shift. Anything which can be automated and is repeatable will be, will be replaced. I think that's that's number one, number two in financial services, and that is like a general, general answer, I would say, is using AI to help people in their daily tasks, by using the services where they can learn, where they can look for the answers, where they can use AI to study topics, where they can use AI to generate, generate ideas, actually, and then just go through those ideas. So I think, like a personal assistant is a great one for any industry, and for Finns especially, that's that's like, I would say the second one, and then possibly the third one for the workforce would be the agents, which are actually used by the workforce to do the daily tasks on behalf of them. So you could, you could easily if you would like to, let's say that you would like

to fly to another city. Then you would ask your agent, who would ask the other agents to do that? And it's and it's tightly connected with the first one, with automation. But I think that using these agents is a is like a third topic, and if I may, I think what is important for the organizations and for any department is just to start thinking about AI today and not to wait for some innovation program or broader initiative. It's really if I'm running a department, the critical question for me would be, what can I do with AI in my department and start educating myself such that I'm ready for the future. And I already do know what is achievable, actually, and started in a bottom up manner, if you if you wish, because I look at AI as I can see it as a for example, let's say Excel, Microsoft Excel, of this century. So in the very beginning, you know, Excel was a new thing. Maybe there was the only guy who could operate it in the company? However, today, everybody

is using that. You don't have a separate unit, which would be like Excel experts, and if you would like to, you know, do a spreadsheet, he would talk to them, and they would do it for you. No, not anymore, right? So, and I do see it so new, new piece of technology, and everybody, and. Is to adapt that, and the sooner you do it, the better.

There's some great points there that, you know, I'm really on that AI train at the moment, because it's not going away, and we've really got to embrace but it's a bit like, you know, what they used to do at Google, where you got 20% of your time effectively, Fridays were given to people building tools, right? You know. So how hackathons? How do you actually, how does the business kind of point mandate or build a culture of using AI? Because I'm hearing some great things that people are building themselves. So let's get to what did AI tell us? So there are three points to the to the response that AI gave us. It's quite interesting, actually, very, again, very, very much aligned to Stepan, which is, which is really important and great news. The first is the way that employees are going to shift. They're going to shift towards more of a higher level of things. So I'll give a bit more description. So you're gonna have to be a higher level of those analytical

skills. It's gonna have to have a higher level of creativity. And also it references another particular role which will be around about relationship management, so really sort of like pushing up that whilst the AI takes on those repetitive roles, as we've described. The other thing I thought was really interesting is that the overall workforce will decline, as you mentioned, slightly. But what will happen is, if you think about an apex the pointy end, our roles are going to start to sort of become a bit more pointed, a bit like to the first point, which I thought was really interesting, hard to describe it. But what I imagine is that we're going to become experts in what we do with by using the tool. Yeah. And then the third thing is that actually what it says is that financial institutions need to get a grip of addressing the work the workforce concerns. So you're gonna have to start to allay fears that people won't be made redundant. They're gonna have to re skill and

change to your point. When electricity came out, you know, people had to skill and learn electricity to use it. So there's the three points.

I think that last point is, is big takeaway for me, Stepan, where you say, if you're a department head. Don't wait for a transformation program to come around. Start to look at how you can, how you can embed this now, because not only will you be potentially left behind in your side, your own organization, but you'll definitely be left behind outside when you look at your competitors, which of course, impacts things like recruitment, bringing the highest performance into your business, all of those other all of those other elements as well.

Absolutely, absolutely. And it also, like you do the thinking in advance such that you can, you can when, when, for example, the you hire a company like ours, like Adastra, we can help you with the topic. However, you've got your ideas already ready. You do know what you are looking for and so on. So that's that's number one. And number two, you can start actually looking at what is already available at the organization. There is a good chance that some somebody else in some other department is already using a tool, and you just might reuse it, or get a new subscription to the license for that stuff like that. So, so I think that you can do a lot by just by yourself, and then then, like from a systematic point of view, there are either internal departments or external consultants who might bring you to another level. However, the basics might absolutely be done by by the department.

Yeah, I think that's it, isn't it? It's the there's so much you can do right now, but it's also, if you want to get that next edge, if you want to think about where you can stand out versus your competition, then that's a really, really great opportunity to bring in experts to help support that absolutely well. Stefan, thank you so much for coming in and joining us on the growth workshop podcast. We've hit so many notes today, and they're really, really important ones, the importance of, well, actually understanding more about your journey and how that's shaped, the way that you are, and building businesses from an engineering background, clearly having that scientific approach on and how you how you tackle that, and how you're, you know, obviously a huge advocate and expert in AI, and I think some really good, helpful guidance in there for for growth leaders out there when they think about things they should be thinking about when it comes to AI, and the impact

it can have on their business, but also from a potential threat perspective, but also a whole heap of opportunity out there as well. So shepan, thank you so much for coming and joining us today. Hopefully we have you back again soon.

Thank you so much, Matt. Thank you so much, Jonny. I really enjoyed talking to you today. I enjoyed this session. Thank you so much.

Thank you Stephan.

Cheers.