Michał: Everyone at a given moment learns that they can now launch a new Product. Everyone wants to offer it to their customers. And now, if I can Offer it in two weeks, and a bank with a significant technological legacy Can do it in two months, then I already have an advantage over them. So we also see the risk that these big tech companies have made us somewhat Dependent on themselves, on their AI, on their models. We won’t be able To move away from them, and there will be a price to pay. 

Bartek: Hi, this is Bartek Śliwa. Welcome to another episode of the Euvic Talks podcast, Where we connect the worlds of business and technology. And today, let’s look at Technology in banking. My guest and yours is Michał Grela. 

Michał: Hi . Nice to meet you. 

Bartek: Michał, first, I’d like to decipher Your position as Head of Growth. What does that mean? What do you do? What was Your business path that led you to this point? 

Michał: That’s a very interesting question. As you said, I deal with the broad Development of Speednet’s businesses. Speednet is a partner In digital transformation in banking. That means we’re working with Banks, fintechs, and insurers to help them Launch new financial products more efficiently and quickly, Modernizing legacy systems—which is often a problem in this industry—without such A big bang, but rather gradually helping them implement innovations, also in Compliance with and the regulations that apply to this sector. At this company, I’m responsible for broadly defined business development. That means, to Put it simply, how much money we earn at the end of the year and what our margins are. It’s actually many components: Sales, marketing, working with existing clients, strategy, a bit of product— So many different threads. Previously, I always connected the worlds of sales and marketing, and Either straddled the fence between sales and marketing or Hovered a bit closer to sales, a bit closer to marketing. Even before that, I had a stint in my legal career. It was a bit too Boring for me, so the dynamics of the technology industry now suit me perfectly . And a role that’s so cross-cutting and connects many Different contexts at many different levels of systems thinking Is something I feel very comfortable with. 

Bartek: Great. You mentioned compliance. Since we’re In banking, and just before we turned on the camera, We were talking about how everything can now be generated, including vision and Voice. So, in the era Of avatar generation based on our image, Yes, our likeness, How do banking systems fare in the face of potential Attacks on KYC processes and user identity verification? 

Michał: Yes, that’s a very fundamental question, a very pertinent one. It seems to me that the Area of ​​fraud and antifraud is one Most exposed to AI in banking, both from The perspective of potential attacks and existing use cases. I think that this area currently has the highest number of such Real-world implementations, return on investment, and profits from AI in banking. And these modern AI systems are very good at dealing with someone Trying to cheat using AI. So we fight AI with AI. 

Bartek: Definitely yes. Definitely yes. 

Michał: Well, the Fraudster, which tries to impersonate someone, is armed with this weapon, but so is the bank. And today, It is here that such tangible benefits And, let’s say, such everyday implementations are most common, and with great Success. This technology is moving fast. 

Bartek: There are new things happening all the time, novelties. LLM providers are competing in providing models. And In terms of banking, I wonder what has caused such A wow, such a boom in recent months. What has changed in these technologies supporting banking? 

Michał: Sure. You know what, that’s a very accurate observation. It seems to me that even We, working in this technology industry, have trouble keeping up with The pace of change, and if I were to keep track of new models… 

Bartek: We keep up. If I were to keep track of new models and all the releases of features from Anthropic, Gemini, OpenAI, and so on, I would probably be the lion’s share. He had to devote part of his week to this. And banks have even more Trouble with this, because, above all, they operate in a heavily Regulated context, where the regulator is closely monitoring their work. This AI can’t be Used for many of the same applications we would use it for. Wherever AI is, it’s still a black box, and it’s unclear Why the machine did what it did. So, we can completely Exclude it from the banking sphere. And generally, it seems to me that this entire Industry is partly at the forefront Of innovation, and partly a bit behind when it comes to its implementation, right? For Example, when it comes to cloud adoption, which is somewhat off-topic to AI , banks can only enter the cloud once the regulator has given their Approval. This has only been relatively recent compared to other industries. When it comes to AI, it certainly won’t be applied in this industry wherever It affects critical systems. This is regulated by a multitude Of different legal acts. The Polish Financial Supervision Authority (KNF) has its own directives. There’s a whole AI law that Says this AI can’t be used, for example, for credit scoring, to Make decisions without a human, as if it were a human-in-the-loop aspect. However, this Doesn’t prevent the industry from adapting to Implementing this innovation in compliance. Hmm. And the biggest news, which is currently on Everyone’s lips, is the entire area of ​​agentic AI, especially agentic AI in the Payment and payment context. I mean, whether I’m an agent or the endgame of the industry, for example, The commercial sector, I can message ChatGPT or any other model I’m talking to, recommend some running shoes, and they’ll recommend the Running shoes. I’ll tell them, “Okay, I’ll buy them And pick it up from the parcel locker.” And the entire transaction could take place Autonomously through agents. This is exactly what Everyone is working on now, to get to this point. The industry isn’t there yet . Will it ever be… maybe to some extent, but I don’t think so. However, it’ssomething everyone is working on now. Banks, Infrastructure providers, Visa, Mastercard, model providers, are all trying To agentify banking. Hmm. And this seems to me to be A good summary of the path that banking is taking In terms of AI, from generative AI Or increasing productivity on the part of delivery teams Through bots, chatbots, voice bots that Answer questions. So, you can go to your Banking app, ask when my deposit is about to expire, and it will answer, but it doesn’t Perform any actions yet. Now banking is at the stage of AI, Which, within a closed scope and under the supervision of a human in the loop, is already performing some Actions, both in the bank’s back office and front-facing. Okay. You touched on the topic of AI in action, And you also said that this technology is already being used In banking in some places, but in others, for many reasons, It isn’t or can’t be used yet. Now I wonder what really poses A greater threat to banks today. Is it the hackers mentioned somewhere here Who will try to steal money, or identity, or data in general, Or the regulations that, on the one hand, are supposed to benefit bank customers, Right? On the other hand, from what I hear, they are also a major limitation in Implementing this technology, right? And how do we maintain a healthy balance here? 

Michał: Yes, there are definitely two sides to every coin. It’s the same here. There’s no Industry conference in the banking sector within the European Union where this Regulatory issue isn’t always at the forefront. Because We see a significant difference in the approach to innovation in the banking sector Within the European Union, and in the United States, or I don’t know, in Asia, where You don’t have such restrictions in the States and Asia when it comes to compliance, there’s a bit More of a push for innovation without perhaps worrying so much about Compliance, legal regulations, and security. In the European Union, Many voices are raised that the sector is definitely overregulated , and that each new regulation doesn’treplace another, but rather It adds another layer of complexity and sophistication. However, at the end of the day, many of these regulations actually Enable innovation. They enable new things, Introduce universal standards, such as, I don’tknow, Open Banking, PSD2, or the upcoming PSR, PSD3, and FIDA. These are regulations that allow For new things across the entire sector And enable new business models and development. When it comes to AI Acts and regulations regardingAI, I tend To be a supporter of these regulations and this approach in the European Union, especially in the banking sector. Perhaps within the AI ​​sector in general, I believe that rapid development and innovation are valuable, but when it comes Toapplying this technology to the banking sector, Security, compliance, and regulations are definitely beneficial to Users. We don’t face the risk, as in the United States, of not being able to talk to someone physically at the bank because You’ll only be… or having to pay to speak To a physical employee. Thanks to GDPR, for example, We’re not in danger of our data being hidden somewhere, who has it, Or to what extent it’s being used. We’re 

Bartek: Protected in that regard. But leaks do happen. 

Michał: Yes, of course, there will always be someone who will try To exploit the system, but personally, when I entrust my data to someone because I’m covered by GDPR, I feel a little safer than if someone were Training models on my data, or if AI were to decide whether I’d get a loan or not. In the European Union, I’m not In danger of that. It’s certainly true that Some of these compliance areas Are a bit confusing and require allocating a certain Budget. I think that within, say, Bank budgets, probably as much as 20-30% of budgets must be allocated to Regulatory compliance, because there have to be dedicated departments dedicated to it: lawyers, compliance, and Audit. Someone is simply monitoring whether what we’redoing is compliant and Secure. At the end of the day, it improves security, and this Usually addresses the issue that poses the greater challenge during Implementation: control, security, and filtering. Is this AI model doing what I want it to do? Is it Responding in line with what I’ve taught it? Is it guiding this customer As it should, is it not breaking the law? For all of this, We need this regulation, and it Allows us to implement this AI safely. 

Bartek: Hmm. Okay. We’ve mentioned hackers and regulation as a potential threat, or at least A certain difficulty in this business. I’m also wondering, because It identifies a third source of threat: so-called neobanks. You know, a Brazilian neobank is already operating strongly somewhere in Latin America , but you don’t have to look for it on that continent. We have Revolut, Which, from what I read today, Will be opening its first branch in Barcelona. And the question is, are the traditional banking system and the banks we know afraid of This new player—not so new anymore, but still a player who seems To have grown on technology and is definitely ahead with it 

Michał: ? How do you look at this issue? Yes, it seems to me that it’s a very interesting Competitive environment, looking at this industry, where we have, let’s Say, incumbent players, i. e., classic banks like Incumbent banks such as PKO BP, a bank that has been Traditional for many years and has grown more by providing Financial services than simply being technology first. We certainly have Fintechs trying to do some areas of banking faster, more efficiently, and Better, and compete with banks here. However, it seems to me that today’s Market consensus is that banks cooperate with these fintechs and, where They can, simply implement these solutions. Instead of reinventing the wheel , they simply partner with such a fintech to Truly expand their service portfolio and do things in a way that’s Simply faster and more efficient. And we have this category of entities You called neobanks, which primarily hold a banking license. Revolut Recently obtained a license in the UK, although, from their Perspective, it hasn’t really changed much. But they’re increasingly entering Traditional banking, for example, in Switzerland recently or soon, Offering private banking services, something that ‘s precisely what one might say is stereotypically Reserved for traditional players. But we also have many European examples like, I don’t know, Monzo, Starling Bank, N26—these are new banks that started recently, Which certainly have a technological advantage, right? Are There also a few such entities in Poland that can simply do things faster? And now it seems to me that they’re targeting Their services to slightly different market segments. Still. You know what, I think the need to go to a bank branch Will never disappear for most banking customers. For many banking customers accustomed to traditional Banking, taking out a mortgage through a banking app Might simply be a step too far. Hmm, maybe for my child, Or for me in 15 years, who is, let’s say, a bit more Tech-savvy, it will be okay, but I think that with such large Amounts, or weight, or some, I don’t know, for example, Corporate situations, working capital loans, business services, that Physical contact with the bank is simply very valuable. The same applies to Private banking. If I’m, I don’t know, someone who Has a lot of resources and needs professional management, I don’t Want to do it through Revolut. I need a private advisor and Someone who takes care of it personally, and a chatbot there, I suspect it won’t Revolutionize. So this market has many different segments. Some of these Segments are susceptible to this, to this neobanking disruption. On the one hand, one might think that Generation Z is entering the market, a digital-first generation that primarily needs Digital-first offerings. They want To handle everything through a mobile app and talk to an agent, a chatbot, and A voicebot. But if we look at demographics, There’s a small number of Gen Zers in Europe compared to the silver generation and older. However, A large portion of banking customers—those who actually have money, and that’s where Banks really make money—deposits and loans—are high-hold Products. It’snot as if the source of these Earnings isn’t Gen Z. It’s more likely to be those in their 50s, 40s, and 60s who don’t want to talk to a voicebot, right? They do need Traditional banking, so it seems to me that this market is Relatively fragmented. In each of these areas, there’s some competitive friction between These three entities: let’s say traditional banks and fintechs, and Everyone can find a piece of the pie. 

Bartek: Okay. If you were to point to one trend or one area where Traditional banks should be inspired by the activities of neobanks, What would it be? 

Michał: I think this area Of ​​technology, especially related to such Technological legacy, is the biggest Difference, right? I mean, if I, As a relatively technologically new bank, , am able to create a new product that appears to everyone at the same Time, for example, a secure 2% loan, right? Everyone at a given moment Learns that they can now introduce a new product to the market. Everyone wants To offer it to their customers, and now if I can offer it in two weeks, and A bank with a significant technological legacy can do it in two Months, then I already have an advantage over them. And shortening this time-to-market for new products is often blocked by Technological legacy that banks have built up over the years. A large bank Still relies on IBM mainframe systems, COBOL. It’s as if they have to deal with this, and they cope by, for example, Adding new products alongside the core systems, Using Some more complex banking, modern solutions. But each addition of such a piece of the puzzle complicates it, and it becomes A real Jenga, which is difficult to understand. The level of familiarity With understanding such a system at such a high level Of complexity is very high. So, this adds another Layer of challenges. So the solution is to modernize Technology. I think the solution to catching up With those leading the pack is definitely The ability to act quickly without being blocked by Technology. However, no bank can make such a big bang and Say, “Okay, we’re changing the core system now, for example, from the mainframe To some modern system,” because it’s too risky. It’s Open-heart surgery. In such a large bank, millions of transactions occur every second , and you can’t simply switch it off. No management board will make such a decision, because the Transition process itself usually exceeds the board’s term of office, so no one wants To start and disrupt the project. That’s why we create these, perhaps Not monsters, but rather Frankensteins, where we have this Core and, for example, new paths, user journeys built around it. We’re already running these new streams, as it were, but we still Have to manage the old one somehow. And here again, AI is perfectly suited for This. Because when we have an old, 30-year-old system with no documentation, understood by Three developers, two of whom have already retired, it’s A significant risk. However, AI today is capable of understanding what’s happening there to Some extent. Of course, it’s getting better every month. Some large providers have dedicated models that can Analyze the operation of a given mainframe or COBOL system, create Documentation for it, and try to modernize it somehow. We’re even Expanding beyond the banking industry, because this is a problem affecting 

Bartek: A broader range of businesses. Who, in your opinion, should be responsible for monitoring this, and how ? Are we really facing the Problem that, damn it, we can’t effectively implement New functionalities or new user Journeys? And on what basis should decisions be made? Okay, We’re modernizing. Okay, we’re changing. Okay, it’s still okay, so maybe we should put that 

Michał: Aside. This is a topic that usually Involves many different stakeholders in banks. There’s always some classic, maybe not conflict, but there are Several actors involved. There’s definitely a business that wants things to be fast, Efficient, and effective. I want a new product because I have P&L target, and I Have to release, I don’t know, a new feature in the mobile app because everyone Can already pay for parking with a mobile app, and we can’t yet, and we ‘re losing customers because of it. Furthermore, they usually have goals related to Increasing the bank’s revenue, selling products, or Acquiring new customers. To acquire new customers, everyone is Essentially competing for a limited pool of customers, every bank, so you have to Differentiate yourself in some way. And business is committed to doing this. Then we have technology, Or the entire IT area within a bank, which must somehow Address these business needs. And now There are many different business stakeholders. They compete for technological resources Within the bank. That’s why banks leverage collaboration with External partners, fintechs, and companies like ours. And they must ensure that this technology doesn’t become a barrier for them, Both in terms of the acceptable level of legacy and the pace Of implementation. But in all of this, there’s also the banking security function, which Is simply security and compliance, which must agree to all of this. Hmm. And in fact, the collaboration of these three, Let’s say, three actors: business, technology, and security. Compliance is something that… this is precisely the place where Such escalation should take place. And ultimately, Such decisions are usually made by the management board, and they reconcile these three 

Bartek: Conflicting parties somewhere. If I understand correctly, it’s a Collaborative effort to recognize the need and ultimately Recommend a proposal to the management board. 

Michał: Yes. You know what, I think every change has its champion. Hmm. And someone who May not be the brakeman, but sees a bit more risk than that champion. Hmm. And it’s always a matter of compromise and discussion. 

Bartek: Okay. We’ve talked about what Banks are working on today. I’m wondering what, in your opinion, is So over the top, so simply no-brainer, that, okay, this AI-based technology suite should already be implemented in every self-respecting bank . What should we expect, and in your opinion, Are banks already there, or do they think they are, but haven’t necessarily 

Michał: Gotten there yet? When it comes to AI adoption, banks Are definitely already there, and this AI is actually saving them money Or allowing them to earn more. It seems to me that The standard that’s already implemented in most banks today Is AI, which is often generative, not yet Agent-based, and is partially directed Outside the bank to the customer, and partially internally. The first Such meaningful use cases were primarily internal, in the bank’s back office. For Example, I can submit an offer to a customer faster because AI helps me Prepare it, and I receive it ready. 

Bartek: So there’s some argument, Right? 

Michał: Yes, yes. Some kind of automation for collecting Information and data from several different places, managing it, consolidating it into one place , and, I don’t know, creating some kind of underwriting process, for example in the Context of insurance or call center optimization. These are Things that are already implemented today, and generating revenue. 

Bartek: Is this Anti-fraud one, I think, one that’s practically standardized, Using AI anti-fraud tools to combat AI Fraud? Hmm. 

Michał: Another aspect of customer-facing Is all kinds of chatbots. I think that today, a growing number of banks Whose clientele expects this have already implemented these chatbots. There’s a very Well-functioning chatbot at Nest Bank in Poland. Vela from Velobank is a Prime example of a chatbot that already does things, 

Bartek: And it’s a chatbot that… What things does it do? 

Michał: Velobank has a very well-organized complaints management system. Hmm. So, for example, you’re a customer, you open the app, you see that the bank Has charged you 5 PLN for card processing, and you say, “Hello, hello, Mr. Bank, why did you charge me this 5 PLN?” And normally, you could Either go to a branch, send an email, write to the chatbot, or Call the bank to explain where the 5 PLN came from. And so the entire process of complaint acceptance, acceptance, and physical processing began , checking whether the 5 PLN charge was applied in accordance with The regulations or not. For example, as a customer, you have to perform Five transactions per month for the card to be free. And someone has toPhysically verify whether you did it or not. And today, it’s fully Automated. That means you can report it to the bank through any channel. A Physical agent, like an AI agent, will accept the complaint and Forward it to the appropriate department, which also has An agent at the front desk who can verify whether you actually had the charge, Whether and on what basis it was charged, and whether the basis Was actually met. So, he can verify that Mr., I don’t Know, Jan Nowak didn’t perform five transactions totaling X PLN Per month, which is why he charged him 5 PLN. The entire process, carried out by Agents and AI, is then transferred to a human, who simply checks and sends a few, and then 

Bartek: Sends them to the customer. Okay. We’ll always have this human in the human-in-the-loop. But will we? Do you think any of the banking roles 

Michał: Will be fully replaced? Humans will definitely always be there, and It’s undesirable for either side to be gone. This means that neither from the bank’s perspective , nor from the customer’s perspective, nor from a regulatory perspective. The lack Of a human in the process isn’t exactly the ultimate solution from a regulatory perspective . Because someone has to check the system, take Responsibility for its operation. Secondly, from the customer’s perspective, you need to have the ultimate option To simply talk to a human. Many customers don’t want to, I don’t know, call A call center, talk to a bot, or use the bank via a chatbot. However, You need that human in the process, A kind of customer-facing process. And from a regulatory perspective, it’s simply Required that these high-risk systems— Most of these banking systems are medium- to high-risk systems— Simply have that human-in-the-loop aspect, so it’s something that’s a 

Bartek: Requirement, I think. I’m still thinking about that example Of a chat agent, a chatbot, that you mentioned, because if all Chatbots were this helpful, I wouldn’t be writing every time I wanted To talk to a human. By the way, it’s a funny fact: just Write GDPR and a human automatically joins in. The question is, do you have any other Examples that were actually useful not only to bank customers but also Allowed the bank itself to increase customer numbers , ROI, or some other metric that is, well, quite 

Michał: Important from the bank’s perspective? Yes, I think this area Of ​​complaints is good. Many of these examples involve 

Bartek: Making it easier for a customer to purchase a new product. Hmm. 

Michał: So, for example, you can write into the chat What deposits you currently offer, and there are already chatbots that can 

Bartek: Open those deposits. Hmm. 

Michał: Which simply simplifies and streamlines the entire process. This area of ​​banking agentization, or the transformation from a chatbot that responds to a chatbot That does, well, that’s just beginning. This year is the year with The highest number of such implementations. Where the real Savings are today are in the KYC area, the Back office area, the customer service area, and the area Related to such things. Well, it’s still customer service, like Complaints and chargebacks, right? This is also an interesting Case. For example, when you pay for something with a card and don’t receive the product or Service, you can ask the bank to actually refund Your money through chargeback Releases, which Visa and Mastercard offer. It’s a whole process where you have to Gather certain documentation, log into a special system (Visa or Mastercard), go through the entire process, and at the end of the day, if you meet the conditions, You’ll get your money back or not. And again, there are entire teams Of people doing just that in banks. Today, you can do it 

Bartek: As an agent. Hmm. 

Michał: I think The call center area offers a significant cost benefit when it comes to implementing AI. Today, however, in most banks, You’re initially contacted by an agent who tries To understand what you’re calling about, Verify you, and only then direct you to someone who’s already Somewhat prepared and handles it as a human. But Voicebots are often the interface Banks rely on today, and these voicebots are capable of handling a given Case submission process fully, and in very marginal cases, They anticipate a fallback to a human. And there’s the difference in the cost Of maintaining this entire call center full of people 24/7, who Are actually talking through some kind of decision tree, replaced by voicebots. So, A lot is happening here, and there are a lot of real implementations And profits for banks, although of course, there’s a limit somewhere In the fact that customers don’t always want to use a voicebot. However, I think that in 5-10 years, it will simply be the gold standard. 

Bartek: Because of everything you’re talking about, and also because of my experience with AI, it outweighs There’s one common denominator about the automation of a large portion Of human work. And how do you look at this? Will Many people Lose their jobs because of this automation, or will those employees who are currently handling Complaints, or who were handling them and are now being handled by agents, simply become More efficient, being moved to more specialized, or More problematic areas. Which way do you think this will go? Because, to sum it up, you don’t have to pay ZUS (Social Insurance Institution) to an agent, So I assume that handling a given process is becoming more Cost-effective for the bank. And thus, I also assume that the bank is Able to handle more cases in a shorter time. Right? So 

Michał: Which way do you think this will go next? It seems to me that this is An answer that will be applicable in a broader sense, not just to banking. 

Bartek: Mhm. 

Michał: And I have a prediction that Both of those things you said will happen, Mhm. That some roles will simply disappear irrevocably, because They’ll be automated, or, I don’t know, 10 people will be left with two or Three, because there won’tbe a need for more, because that’s how efficiency will increase. And some roles will change dramatically precisely because the type of work Will change. I see this firsthand In my own organization, which is simply a technology organization Focusing on… software development, and I see that today A team of five people can do what would have previously required 15 people. Does that mean I’ll be left with two of those five people? Probably not. They ‘ll simply work differently and do different things. This also Requires a different competency profile from these people, because suddenly it turns out that The best programming language is English, and The barrier to entry into the industry is getting much higher, right? Well, you don’t really need juniors anymore , because one competent senior with extensive Experience and understanding of the business can do what Previously required an entire team. However, this is a trap, because soon Seniors will also retire. Nobody wants juniors. There’s another risk Associated with this AI, well, you don’t pay it social security contributions, but it also has 

Bartek: Costs. 

Michał: Yes, Infrastructure costs, energy costs, and above all, token costs. These are Things you have no control over. And recently, I think this week or last, We’ve seen several such major changes in the market , when, for example, token pricing changed drastically, and It turned out that if you have a Claude Code plan and had a certain number of tokens, it Suddenly turns out that now it’s not enough at all. Suddenly, I imagine That, for example, due to the costs of maintaining a data center, developing models, And the fact that these companies will eventually have to become profitable, Raising the price of using this AI might not be such a Cost-effective solution. The cost-effectiveness might still be within the realm of efficiency, But financially, it was cheap, it’s moderately cheap, and what it will soon be, I think it’s increasingly expensive. So we also see the risk That these big tech companies have made us somewhat dependent on themselves, on their AI, on Their models, and we Wo N’t be able to move away from them, and the price will have to Be Paid . I Think There’s Some kind of risk of vendor lock-in here , It ‘s There 

Bartek: With Claude in a few minutes, right? And there’s a change. Michal, I finally have a question for you. Not from banking anymore. Or maybe From banking? It depends on your answer. What 

Michał: Has inspired you recently in life? It could be… What has inspired me in life? 

Bartek: In life, Yes. It could be an event, a book, a movie, An experience, a trip, a journey. Oh my. 

Michał: I think that professionally, This area of ​​agentics is one that strongly Inspires me to try to be the first there And to actually do something there that will make a difference. Personally and personally, I’m expecting twins. 

Bartek: Oh, congratulations! 

Michał: I already have one son, I’m expecting twins, and it’s allowed me To look at many things from a different perspective . And it also shows me a little bit about what is truly important, and I think it’s given me such great Inner peace, because that’s all I Really need right now. Just to have that peace of mind, and I see it Spilling over into all the other areas as well. 

Bartek: Okay, great. Industry sarcasm. Since you’re expecting Twins, is Gemini your favorite model? 

Michał: Oh gosh, I guess there’s no other choice now, but 

Bartek: I actually use Gemini the most on a daily basis. Great, 

Michał: Great. Thank you so much for the conversation. Thank you. 

Bartek: Thank you for listening, watching, and I invite you to another episode of the Euvic Talks podcast