Felipe Argento, co-founder of the Cartesi Project, welcome. Thank you. Thank you for having me. I spoke with Jill Gunter from Espresso about two weeks ago and during our call she mentioned Cartesi. That's when I looked up Cartesi and I thought to myself, wow, I need to speak with these guys and get them on the pods. I'm so glad that you said yes and that you're on. That's great feedback to hear. We love the Espresso people. We have such a nice friendship. We do events together. do some code integrations. They're so nice. that's good. That's good. Well, we'd love to hear about Cartesi and what you're doing. What I want to start out with though is kind of your origin story. I know you're a developer. You've been a developer for a while. I'm curious to know how you got into crypto and also what the crypto culture and how the crypto community is like in Brazil where you're from. Sure, So I actually stumbled upon crypto. I was on Reddit, you know, just kind of stumbled upon it. And then I found I was at a college during that and I found a study group that was looking into crypto and so on. So I joined that group and eventually I started my own company. which is not Cartesi, it was a development company and a consulting company. At the time some big companies in Brazil like electricity companies and so on were trying to do something with blockchain. Though it was at the very beginning, you know, the era in which we would try to decentralize everything, the ICO boom and so on. Then I started that project and then one day Eric Demora, which is another founder of Cartesi, another co-founder, he called me and was like, hey, I'm starting this project, come join me. And I was like, no, I have my own thing, why would I join you? And then I went to look at the actual team that was building Cartesi with him. And then I found two math professors at a... at a math institute in Brazil which is one of the most famous math institutes, mathematics institutes in Latin America. There was this guy in Taiwan and I'm like okay these guys are much more prepared to face the world than I am alone. Then I called him back and like I'm sorry actually I went in and that's how I started. But at the beginning I fell in love with the It makes the three things that I really like, which is like economics, politics, and development. So this like game theory aspect of it. And on the Brazilian community, you know, I feel like us in Latin America, we have a sort of a different relationship with crypto because, well, when crypto first came about with Bitcoin and those... projects, there's a lot of talk about inflation and protecting yourself from state currency and so on, from manipulation and those kind of things. And in Brazil, well, in Latin America in general, but Brazil specifically, we have dealt a lot with... currency problems here, right? this idea that yeah, money getting devalued. And I eventually moved on from the monetary aspects of it. The part I like more about blockchain now is human coordination and these other kinds of challenges. But the community in Brazil is quite strong because of this complicated past of government trust. That's cool. I've worked with a number of developers and also founders from Brazil. yeah, I mean, my sense is that it's growing and it's very strong. And it's pretty cool that DevCon is gonna be in Argentina this year. So that's pretty exciting. Yeah, DevConnect. Yeah. So actually, there's a funny story. So in Brazil in the 90s, we had like an insane inflation. prices at a supermarket were getting updated daily, right? Even sometimes hourly. And as a result of that, our banking system was always one of the most advanced banking systems in the world. Because if I send you 5 reais today and it just arrives to you tomorrow, at that time it would be worth half the amount of money, right? Because it would actually go down daily. And nowadays our central bank is actually one of the most advanced central banks in... in the relationship with blockchain, we have a lot of government projects using blockchain for transparency, digital money and so on. It's kind of cool. that. That's really, really cool. Let's talk about Cartesi. Maybe explain to us kind of the problem that you saw that led to the development of Cartesi. And then let's get into Cartesi and how it solves that problem. Sure. So the idea behind Cartesi is that, so we're trying to do as an industry, we're trying to do decentralized compute, right? That's basically what we're trying to do. And then it turns out if you looked at blockchain, the industry in general, five years ago, and also if you look at it now, we are losing the computing part. So we did a very good job at having the decentralized aspect of it, but the computing, in the sense of computing powers, scalability, expressivity, all of that was lost. Which turns out, in turn, we only have simple applications to play around with. We have a very limited environment to do the computing. And that's what we tried to solve, right? So we focused since the beginning in creating an execution environment that was scalable, that was powerful, that was expressive. And now five years later, six years later, we have... what I think, I could argue this with anyone, it's the best execution environment for blockchain. And I can explain why I can safely say it's the best. But that was the idea, was to get computing back to decentralized computing. So we can have real applications, not real is a bad word, but so we can have complex, intricate applications in Web3 as well, and not be only limited by very simple things. got it. Now from a technical perspective, it's, so I'm reading from the website, Linux powered rollups, the simple way to build in Web3. Tell us about the Linux part and why that's important. Sure. So this is a little bit on the concept of bringing the computing back. The entire internet was built for Linux, right? Everything that you know, all applications that you run, they assume that under what they're running, there's an operating system, usually Linux. So when we use libraries, coding in an important library, that library assumes that there's a file system and a memory management system and so on and so forth. So the idea is that by having Linux running on your execution environment, that extends to you the capacity to use software that was made for Web2. So when I talk about Cartesi, I like to use the concept of backwards compatibility and forward compatibility. So what we're trying to do is to bridge all the code, open source code that was written for Web2 and bridge that into Web3. And Linux is what makes that possible. So you can see Cartesi as a bridge from the... software tools and infrastructure for web 2 to web 3. And it's backward compatible because you can use what was built in the past, can use Rust, Python, NumPy, everything that runs on Rust 5, but it's also forward compatible, right? So recently we had the whole DeepSeq launch that crashed the markets for a little bit. And we actually have an application on Cartesi that runs DeepSeq. Because of course, DeepSeq was built in a way to make it simpler. You have Ljama, which allows you to run an open source AI model. And you can run Ljama on Cartesi, right? And therefore, you can run an application that's trustless on Web3, connects to Ethereum, or connects to any other EVM compatible chain or any EVML tool. And you can run DeepSync on that. Of course, DeepSync is super slow. It's not yet feasible to actually run an application that runs DeepSync. But you can get to a few tokens per second or something like that and actually run it on chain. Sorry, I'm late. Yeah. No, that's helpful. The sub-headline says, Kartezi is a powerful modular blockchain protocol that equips developers with access to a full Linux environment and high-performance roll-ups designed to support next-generation apps. And so you're positioning Kartezi as a modular blockchain. So it's not an L2. It's an L1 on which L2s could be built. Is that correct? So that is a bit confusing and it's good to have this feedback. What we mean by modular blockchain protocol is that there are a lot of problems to be solved in the industry. You have data availability problems, have consensus problems, have execution problems, have UX problems, liquidity problems and so on. We're focused on solving the execution problem. So we are an execution environment and it's a modular execution environment because you can run it in many different ways. So you can run it as an L2 on top of a theorem, you can run it as an L3 on top of say zk-sync or arbitrary. Or you could run it as a layer one, but that requires a lot of other pieces such as consensus and so on. But we usually run Cartesi as an app chain rollup or an app chain layer tree. And we also have a co-processors going on. It's a new initiative that we started. So Cartesi co-processors on top of IEA layer. So that's why we use the word modular. But you should think of Cartesi as an execution environment that can be plugged in into many different protocols. I see. is Cartesi, so when you think of these like roll up as a service providers, RAS providers, is Cartesi one of the options if I wanted to build an app chain or a roll up, could I choose to build it on, on Cartesi using one of these RAS providers? I don't think we're integrated with REST providers yet, but you could launch a Cartesi app chain and you just deploy it and there's a service provider called Sunodo that is Cartesi specific and it can run your validator node for you and so on if you don't want to run infrastructure yourself. But you also have what I believe is the main setup at least for the very paranoid users, which is that you are always able to run the application that you're interacting with on your machine. So you can secure, make sure that it's running securely. there are a lot of projects that are built on top of Cartesi. And I want to talk about these in a second, but it's very impressive. Can we talk about like the target customer for Cartesi? I imagine it's current developers, and please correct me if I'm wrong, it's developers that are trying to build a roll-up or their own chain, and then they wanna use kind of for the lack of better terms, like a easy plug and play kind of software development kit underneath, something like that. Yeah, that's a good question. So yes, the Cartesi is used, is an SDK that is used by people who are trying to build their own chain or their own application, right? Because applications can be changed in the AppChain model. So it's aimed at those who want to build something that is not buildable on Solidity, right? Solidity imposes some limitations. You cannot import decades of existing software and libraries and so on. So we used this example of... Angry Birds, right? If you wanted to build Angry Birds on chain, if you do it in Solidity, then you'll have to the gravitational loss of physics in Solidity, right? If you are to code Angry Birds on Carpeze, you'll just import a physics library that exists in basically every language that you have, At the beginning, we were focusing on attracting Web2. founders to work on Web3. So the idea that since you have all these suits of software kits that you import from Web2, the idea that you can use that on Web3, we thought it would attract people from Web2 to build on Web3. Then we bumped into a problem, which was that we ended up having to explain what Web3 was to these Web2 people. Hmm. And then we pivoted a little bit to try to attract Web3 people and convince them like, hey, look at the amount of tools that you have if you use Cartesia. You can use TensorFlow. You can code in C, C++, REST, Python. You can import mathematical libraries and whatnot. And now I think we are a little bit back into bringing Web2 developers, because Web3 is more well-known. There's more interest. And since we can easily onboard all these developers that are used to working with Web2 stacks to Web3, then we have a very good target audience there. Are you looking at, when you think about your target audience, you looking at specific developer communities like Rust developers, Python, and then maybe doing outreach and messaging to those specific language communities? Yeah, we went less by language and more by vertical, right? So we got in contact with some game developers, for instance, that were used to building games for Web2 and now they're building games on Cartesi. We didn't go specifically for targeting a programming language group, but that's an idea. I bring that up because I'm, I was looking at, I'm actually really surprised. I'm not on LinkedIn very much, but today I was on LinkedIn looking at groups and I noticed that there's a very big solidity group on LinkedIn. There's also of course Rust and Python, you know, Matlab, like lots of these language groups on LinkedIn. And I was actually really surprised. And I thought, wonder if anyone's actually targeted these groups to, you know, the, maybe protocols that have, that could, that need developers to build on them to, um, if they've actually targeted these groups. And so I, that's why I brought it up because that could be an interesting kind of onboarding kind of developer acquisition strategy. Um, because most crypto people I know, yeah, most crypto people I know are not on LinkedIn. And so I thought if they're not a LinkedIn, then it might be a good place to actually do some work. Yeah, that's quite interesting. I'll suggest that to our BD people, see if they can do some outreach. I'll get back to you. Yeah. So the question of like on which chain should I build my application is kind of a hard one because there's a lot of options. And so how does Cartesi stand out from the crowd so that it attracts the right developer that it wants? So we tend to say like if you want to build a simple DeFi application and Solidity is enough for you then like use Ethereum, use Arbitrum Optimism, something like that. you want to build a complex application that takes care of an entire virtual, so if you want to do an exchange with order books and landing and risk analysis, then you have to use something more complex that allows you to extend abstractions that is part it. If you want to write a complicated game that involves physics or a first-person shooter or a strategy game that has complex logic an expensive computation, then you can use cartelity. If you want to build something that leverages artificial intelligence, I mean, maybe you won't be able to use the current large language models with 700 billion parameters, but you can use sensor flow and you can use simpler models and stuff like that, then use Cartesian, right? Because of this ability of us bridging trustlessly the Web2 world to Web3, if you want to build Web2 applications with the cool features of Web3, if you want to build Web2 applications that are trustless, permissionless, credibly neutral and so on, then use Cartesian. That's how we think about it. And that makes sense to me. If I were a developer and I heard that, that would actually resonate. What about the challenge around latency and speed? And on which, I guess, if I were to build, let's say, game, one of the key aspects is immediate feedback to the user, to the player. How does Cartesi deal with requirements like that? Yeah, So, a Carpezie app chain. A Cartesian App Chain can answer you about its state super quickly. It has two limitations though. The speed in which data comes in and the speed in which data comes out. Meaning that if you want to make, for instance, Ethereum be aware of the state of your game or if you want to withdraw assets from the game you're playing, then there's a delay there. And there's a delay in which information can come in. So to make it trustless, the data has to be available on a blockchain. So if you're using Ethereum, pure Ethereum, then the inputs to your game, they have the latency of the inputs to Ethereum. But if you're running a Cartesi application on top of base, for instance, which we have some... you can play around, can send you some links later, then the delay for your inputs to cadena is the delay of inputs making it into the bass layer 2. And that's why we're interacting with Carpe with Espresso by the way. Because if you run an application that uses Espresso to receive the interactions to the game and Cartesian to process them, then you can get your game to go as fast as the confirmation layer on Espresso goes, which is quite fast. So it kind of depends on which setup you're running Cartesian on. Yeah, yeah. And that makes a lot of sense to me. Because oftentimes when I speak with founders of either layer ones or roll-ups, they talk quite a bit about what they have control over, but there's bottlenecks that are not in their control. And that often dictates also the types of applications that can be built on them. And so if I were a game developer, then and I want to you and I'm building in a language that Cartesi supports and if you know building on top of Ethereum layer one makes no sense for me and Cartesi does make sense but one of the requirements is I need you know I need immediate feedback on the state of you know this first person shooter game then what that means then is Cartesi plus base might be the best kind of combination I guess Exactly. And if you need a lot of data for your game, then Cartesi plus like Celestia are available or another data availability layer. That's what we mean by modular in the website. But a cool anecdote just to illustrate this is that our first games that we launched on Ethereum were all turn-based games. Because turn-based game is something that doesn't require the fast interactions that you need. You can send your transaction, it's your turn so you can wait for your transaction to make it into a Ethereum block at mind and then you get your response and then the other player plays and so on. When you go to base then you can have a game that handles faster interactions. Maybe not real time yet but I feel like we're approaching that as we go. And that's pretty cool. And so it also dictates the, from a business development perspective, like which category of games that you're going after first, like what. And I brought up the roll up as a service provider, the RAS aspect earlier, because as you're describing the offering of Cartesi, and because it is modular, you know, at some point, these modular kind of services, could get rolled up into a package. And so, for example, if I'm a developer and my game requires XYZ kind of, you know, I have specific requirements, it would be really helpful for me to know, okay, I have these requirements, which means this package of Cartesi plus XYZ makes the most sense. know, Cartesi plus base plus Celestia or, oh, and I also need, you know, a ZK verifier or ZK prover. So, you know, bass plus Cartesi plus Celestia plus, I don't know, Succinct maybe. Something like that. Do you envision a world where there's kind of verticalization of Cartesi being one of the key kind of pillars, but then underneath it is like a package of things, including maybe even espresso? Yes, yes, for sure. I think that's the vision we're going for. There are some concerns on security, for instance, because I imagine that the user will be abstracted away from all these decisions. So you do have to understand what are the trust assumptions that are getting added by each new block that you have, right? So if you're using Cartesian top of Ethereum, then the trust assumptions are very well defined. If you're using Cartesian top of, say, another proof of stake chain that the stake is not well distributed and connecting to yet another data availability committee that you don't really know the trust assumptions on, then the risk is not the same as you're using a more vanilla vanilla setup. So I think the challenge there is communicating to the users which trust assumptions they are adhering to depending on the package that they're choosing. Hmm. You mentioned rather than targeting developers by language, you're looking at categories and you mentioned gaming as one of those. Are there other categories that are kind of a good fit for Cartesi that you're focused on now? Yeah, for sure. So we have a very cool project that I'm excited about, which is a decentralized exchange, but they do dollar cost averaging, right? So you can state something like, I want to buy $1,000 in Ether across one month. And then each second it's going to buy a little bit of Ether for the entire month. like it's played through the entire month. And I always think that that's a well it's such a common strategy in the real like thread file and we cannot do that on Web3 because that's too expensive right it's computationally expensive. and it's a lot of interactions. But on Cartes you can actually do it in a super clever way. And I imagine how much money people would have saved in bear markets, right? Because you can say like, okay, I'm going to sell 10 % of my portfolio in one year and you don't have to actually be looking at the market, trying to time the market. So I'm excited about that. There's another application that I like a lot, which is a totally different kind of application. It's called Bug Buster is a decentralized bug bounty program. What to do on this application is that, well, the developer uploads a code, right? So I upload my software and I say, this software cannot break in this way, right? So say I upload a bank and say, no one can withdraw money from this bank but Alice. And that's there on a Cartesi machine, on a Cartesi app chain. Then if anyone, any developer, sends a command to this program, to my program, that breaks the rule that I set, they automatically get paid, right? So it's a decentralized but bounded program because you'll say like this is a program and this is what cannot happen. If someone forces it to happen they automatically get paid without any... Judiciation, mediation, you don't have to trust that the guy that wrote the program is gonna pay you. Because sometimes what happens is that you denounce a bug and then the guys are like, okay, let me just check. And then they fix the bug and they don't give you the money. And they dispute that you actually found the bug. It's a very cool example. no, both of those sound really cool. The dollar cost averaging deck sounds really amazing. Like I could see that really, really getting a lot of traction. And I mean, just imagine how much money and also time people would save not having to be, you know, watching charts all day, you know, just set it and forget it. Yeah. and they're even adding, last time I checked, they were adding conditional rules as well. So you can do complex stuff, like buy $1,000 of ETH if ETH price is below X and sell$1,000 of ETH if ETH price is above Y. And it's automatically trading. Super cool. really cool. So let's maybe extend that example. So let's say that application, that dex goes live. I imagine that dex will have its own token. How does Cartesi benefit from an application like that? Yeah, so this utility for CTSI comes from a proposal called Validator Marketplace. So the idea is the following. You have a few options when interacting with a Cartesi application. Let's say I'm a user of that text. I have a few options. I can either run. the decks myself and verify just as someone would run an Ethereum node, a full node and be sure that everything is happening correctly or I can trust someone to run it. Now all I need is a single honest person out of everyone that's running I need a single honest person to ensure the correct result. So the valid intermarketer plays is a network of professional software operators that run that run these applications, right? So I can trust a professional software operator to verify that these text that I'm using is working correctly. And that's great, because maybe I run the node myself, but maybe I want to go on a trip and I want to check the information on my cell phone. And then I can connect to this validator that I trust and use it there. And these validators they have according to the proposal, they have to stake the assign to be able to validate. they get fees on the usage of the applications. So in a sense there's a connection between the size of the Cartel's ecosystem, the amount of dives, the demands, and the utility of the CTSI token. Got it. So for the audience, CTSI is the token for Cartesi. Got it. What about settlement? Do transactions settle on Cartesi or is that on, since Cartesi is at the execution layer, then the settlement layer would be Ethereum. Is that correct? Correct, yeah. I mean, as it's modular, can choose the settlement layer that you want. We use Ethereum. Yeah, I think every app that I know is either settling on Ethereum 1 or on a layer 2 on top of Ethereum. So you could use Bayes as your settlement layer. I see. Could it, and Cartesi's only EVM, and so meaning I couldn't settle in Solano. So that's a good question. So yes, all our packages and code is EVM and we only coded on the fraud proof game for EVM but there's nothing fundamental that stops the Cartesi machine and Cartesi robots to also work in Solana. We just haven't built that code yet. I see, or even, or even. to a data as a data availability layer or connect Solana as a settlement layer, you just have to create bridges to cartelism that we haven't. or even bitcoin is that correct? Bitcoin is a bit harder because part of our fraud proof game is you being able to execute a very very tiny step of the entire execution. To prove that what's running on the cartesian machine is running correctly, in the case you get attacked, you might have to prove like an addition or a multiplication or a load on chain. And on Bitcoin you like the programmability of doing that. But having said that, with the BTVM kind of initiative, maybe that's going to be possible. Yeah, because I met with a project just this week that's actually, they're using Solana as the execution layer, but Ethereum as the settlement layer. And so here we're seeing it the opposite. Here it's, know, Cartesi as the execution layer with Ethereum as the settlement layer. And so I thought, well, what's stopping Cartesi from doing the same thing, you know, with Cartesi? and Solana as the settlement layer and the opportunities that could potentially provide. Yeah, that's very interesting. And by the way, I would invite your audience. We have an open Discord, so everything we do, both research and development, is public on our Discord. Everything's open source. So I'd love to get an initiative going to port Cartesi to Solana. And we would for sure help that out if you come to our Discord, if that's something that you're interested in. the show up there and let's talk. I think that could be very interesting. But we do have to focus, right? I think we have a problem. I think we created something that is, I'm in love with it, right? But it's so amazing that we can see so many different uses for it. But in reality, to create a production level code that can hold a lot of money, you have to be very careful. You have to do it in the correct way. You have to verify that everything you're doing is to the... at most standard, you So that limits the amount of things you can do at the same time. So that's why we're focusing on Ethereum. I'm also a lover of Ethereum. I'm in love with that project. So I do have a personal preference for it. But I think it would be very cool to have Cartesi running on Solana. Yeah. I looked at the YouTube page and there's over almost 300 videos, which tells me that, know, Cartesi cares a lot about, you know, developer education and just education in general for the community. Tell us more about the, I guess, the strategy behind that, like why educating developers and educating kind of community members is like, how important is that to Cartesi? Sure. So I think we have a problem with Web3 that we try to ignore sometimes and we try to not talk about it, but it's the lack of applications that are actually being used by end users. So people are actually going on chain and doing something other than just trying to predict what's going to be used in the future. And I believe that the only way to make that happen is by quantity, right? To have the biggest amount of possible people building something that they believe will be useful. And in the numbers you'll get the gems. And the gems are gonna be the apps that are actually used and that people can interact with and it's gonna improve someone's life. So that's what we're trying to do with CartelZ. We built this super cool infrastructure for you to build your own application. But now we have to attract people to actually go and build their applications, So we have to go out there and educate people, show them how to use CartelZ, show them what they can build, and try to find the use cases as an industry. I think that's really good that you're doing your part in helping to educate the developers, the community, and YouTube is a really good way to do that. What else in your content strategy that you're implementing to help bring education, not just to Cartesi, but really to blockchain in general? I guess, how do you think about content strategy that way? Yeah, that's a good question. think, well, we tried to have a dive journey that is very well structured and that actually teaches you as you build. So to guide people through all the steps to building a characterized application. And in the process, you end up explaining things about the base layer, what the set-link means, what are tokens, assets, composabilities. are all concepts that are baked in documentation and that journey. We try to explain that Cartesi is one piece of the puzzle of our larger infrastructure. We are a very important piece, which is execution. But we are solving one problem and there are other problems that are solved by other projects that people should be aware of as well. And we also try to generate content. So we have this very cool event that happens, I guess, once every six months called the Experiment Week. which we stopped everything we're doing for one week and we built something cool on top of Cartelsi. So the core developers were there all day looking at bits and bytes, they stopped what they're doing and they built a game, or they built a crazy DeFi experiment. This last time we partnered up with EigenLayer to do that. And man, the results were so cool. And it's both a way to experiment and improve what we're building, but it's also a way to point to the use cases that we want people to build and try out. So, yeah. Yeah. I meant to comment too on the dev journey. You know, you go to the homepage and the first button there is start building. And then the second page is, you know, three ways to getting started on Cartesi. Like the dev journey is very clean. It's, know, go to docs, read guide, join discord. And I love that because you're, you're, you're virtually hand holding the developer. You know, lot of some developers don't need this, but I think most do. think most are actually like, they don't know what to do. so helping them holding their hand through the process is, so helpful. And so that's, you know, kudos to the, to your, to your level of like emphasis on the dev journey. And I love that. And, people who are going to be very happy with that feedback because there's a lot of hard work in that. It is hard work. It is hard work. And in crypto, we kind of assume that, you know, developers just, they can just do the research on their own. But the truth is a lot of developers, they still need that kind of guidance and coaching and mentoring. And if you can't be there in person, then make sure your docs are clean and helpful and easy to understand. You can find things. No one wants to hunt for stuff. and having it readily available helps so much. We want you, like you come to build on Cartesi, we want you to focus on what you're building, right? And not trying to learn how Cartesi works, how blockchain works, how Solidity works. We want you to sit down and code and solve the problem that you want to solve, right? And let the problems that you're solving for us to solve and you just use what we already solved. And I think otherwise, like... I mean, it's already hard to build a cool application, right? If you have to struggle to find out how to use the framework and read their source code and so on, you're just wasting time and time that could be used for writing cool stuff. No, fully agree. One last thing before we get off that I want to kind of give you guys a compliment on. You know, on the developers page, cartesi.io slash developers, towards the bottom shows meet the teams, building and innovating on Cartesi. And so seeing people's faces makes a big difference. And you know, there's all levels of developers. There's experience to brand new. And what it shows is that, know, Cartesi is for everybody. You know, it doesn't matter what kind of developer you are, what language or level of experience, you know, come, come to Cartesi. And that speaks a lot. So again, that's, that's a compliment for you and the team. Um, and these are YouTube videos. And so these are testimonials again. And so it's, you know, again, when you hear people's voices, when you see their face, it's like, okay, I'm comfortable giving Cartesi a try. And that's really kind of the next, that's really the call to action there. so compliments to you and the team on that. my- Well, Felipe, thank you so much for taking the time to meet with us and to tell us about Cartesi and the great things that you and the team are doing. I will be sure to share URLs and also any links to the project. Any last words you'd like to leave with the audience before we get off? So it was an immense pleasure to meet you and to be on this interview. Thank you so much on this conversation, not interview. Thank you so much for the kind words. And I'll just invite everyone to come to our Discord. We do accept every developer. We're always there chatting, having fun, building cool stuff. So come, you'll learn a lot. And I'm excited to meet you all. Thank you. Thank you, Felipe.