ETHConf Crypto Startup Showdown

Transcript

[applause] Hey everybody, welcome to our startup pitchfest. I'm Cartik, one of the co-founders of ETH global and we're going to get started. So we are here to do an incredible thing which is we're going to showcase a few incredible startups. There are pitching here for our incredible judges. We're going to go into that. There's 14 teams and we're going to just jump right into the program. This is our showdown. We made an open application for past ETH global projects that are turning into companies from the hackathon and a handful of open applications for just any startup that wanted to come in and get a lot of feedback from people attending this

event and potentially get anywhere from funding to business opportunities to partnership opportunities from the crypto and the institutional side. We're going to have three incredible judges come here and give feedback to these teams. Every team is going to get three minutes to pitch what they're doing. Then there's a four-minute Q&A. And for this, I'd like to thank our three incredible judges. We have Tun Chetra, who is the founder, managing partner, Robot Ventures. We have Joyce Yang, who is the partner at Pivot Global, and David Shangard, who runs the Cointelegraph accelerator program. All three of them are early stage VCs. They will help anywhere from give you guidance to what you can do here to uh give you feedback to take this to the next level. And if they're interested,

we'll also talk to you about potential fundraising. And the way we're going to set this up is that this is going to be a mix of feedback and a prize. So the best pitch here does win $5,000 cash. And any project here will have the opportunity to get a lot of space for a future ETH Global event. So you get to be part of any future e- global hackathon for the next year in any part of the world. Uh, same thing for our Pragma conference and we're also going to make you a confirmed sponsor for the next year's ETHCOM which officially confirms there will be a ETH comp next year in New York. And uh these teams can come on and we will make sure that they have the ability to come in and uh actually showcase what they've been up to. And on top of that,

Cointelegraph um has agreed to graciously offer $25 credit for any of the service that they provide to help get your startup out in the world. and get any help you need with distribution. So we have 14 incredible teams. We're going to have Amplify, ASI, Ascend, Cambria Dolphins Frames Frankcoin Glider Paper Pulse Pay Radius Unlink, Heseker, and Sci-Fi Pitch. We're going to just kick things off right away. So, let's welcome our very first startup for this pitch fest, John from Amplify. Please give every round of applause.

Good morning everyone. Oh, the timer is already down. Good morning everyone. My name is John and I'm going to be talking about permissionless prime brokerage. Prediction markets and perexes have been growing at a tremendous rate. However, these markets don't have access to prime brokerage services which is widely used by institutional traders for capital efficiency. When I talk about prime brokerage, I mean credit for traders. Today, only hyperlquid is widely covered by prime brokerage services. And even if they were to cover all markets, prime brokers have a huge onboarding friction since we don't know how to do credit in DeFi. to on board you have to go through

weeks long of KYC and paperwork and this is a headache for traders and this is just not possible for agents. We're building amplify which is a single click access to permissionless prime brokerage to solve this very issue. We allow you to take leverage on any exchange or we allow you to collateralize your position say on hyperlquid to go open a delta neutral position on lighter. The way we do this is by introducing trusted execution environment enabled programmable custody. We allow a trader to post any type of collateral. This could be any D5 position to a smart wallet that's controlled by the trader and the trusted execution environment. The trusted

execution environment verifiable codes enforces restrictions on how the borrow funds can be used and enforc a liquidation condition that will make the lender full. With these protections in place, traders can borrow up to 10 times of their collateral. We're we're talking about prime brokerage, but if we take a step back, this is about permissionless credit, which I would argue is the biggest unsolved problem in DeFi. And this is the right time to build this. We all know how fast these markets are growing and prime brokerage alone today offers a $3 billion revenue potential.

Second is agentic trading is on the rise. The cost of building strategies is going down and the bottleneck will be access to capital. We launched a leverage tool on poly market two weeks ago. Um and we have had over $320,000 in volume and uh we're the right team to build this. My co-founder Orest was an infra engineer at scroll and he has a startup exit in his past. I built the first trusted execution environment based project in our space secret network which did private smart contracts. I built a DeFi system around it and I've been using DeFi as long as Ether Delta was around.

We're raising to fund our expansion into new venues and ship cross venue collateralization. Thank you for your time. >> Uh thanks, John. We're going to go into our first Q&A session. So, we have four minutes for our judges to ask you any questions about the company, your team, and kind of the opportunity, and we'll uh kick things right off. >> Um hey, uh thanks for, uh presenting that. obviously a lot of changes over time from the the kind of uh initial TE space. So I'm curious what's sort of the overhead in terms of like latency for a user uh you know what what are kind of

what do I think about like the median behavior versus worst case behavior you know how have you kind of tested that? So when we offer leverage, we have to um what we do is we allow the trader to borrow um outside of the trading venue and then uh they execute the order on the trading venue's order book with the borrowed funds. So for Poly Market, it takes two seconds for you to borrow on Polygon because that's the um um that's the block time. And then you know you can interact with polygon's I'm sorry poly markets centralized clove API which is uh which is fast. Um so I guess the the bottleneck here is if you need to

borrow that requires one block. >> And who are the suppliers on the supply side for for this? >> Um currently we have a few deals with hedge funds. So, we have um line of credit up to $5 million. And our goal is to prove that um lending income plus liquidation penalty minus bad debt is still a positive APY such that we can open it up to any USDC or stable coin holder. And the idea here is every uh trading pool, every market corresponds to a single lending pool similar to morpho such that the um the risk can be contained a >> and what are the interest rates you're like what how how steep is the

utilization curve? How are you thinking about the quoting? >> So right now the three markets that have the most demand is geopolitics. People bet on um u US Iran permanent peace deal with I think fiveday expiry. Um, I want to say the interest rates around 200% because the borrowing time is so short. We're going to be launching World Cup games where we're going to have a,000% APY because you only borrow for 90 minutes. >> Yeah. >> And we have daily Bitcoin strike price markets where people uh have been uh trading a lot and that ranges anywhere between like 80 to 100% based on utilization of the pool.

>> Cool. I'll let other people ask questions there. >> Yes, man. Hi u thank you very much for the presentation first of all just a little suggestion on the naming I know a few projects that went down I know I'm from a media team as well so uh losing a a a Google search is is not an option so just a little suggestion there so I have a question here uh which is how do you control slippage during liquidations uh great question so it depends on the venue that we interact with if we're interacting with a venue like hyperlquid it's easy because you can just put a stop-loss order as soon as you open the position and Hyperliquid's infrastructure is good enough to to execute. When you're dealing with uh Poly Market, that's a different story.

They don't allow stop-loss orders. So, we're running our own risk engines that have different liquidation parameters for different pools. And um we monitor the order book and trigger liquidations. >> We can uh from detecting a signal to executing an order, we can do that in less than 200 milliseconds. But there are cases let's say in live soccer games when a team scores the price jumps so high that there is a bad debt and and that's why what we're not suggesting that there will not be any bad debt but we're saying with the interest income plus the liquidation penalty over time we should be able to make money for the lenders.

>> Okay. I believe you already answered on who provides the liquidity based on on the previous question but what kind of what kind of assets are actually allowed uh would you accept today we allow only USDC deposits um but the goal is uh and that's your collateral we aim to have in three months um any D5 position any in any supported venue as collateral so that could be your hyperlid bit Bitcoin long that could be your um I don't know lighter position poly market position maybe ki you name it >> yeah and just the last question on the team so you have previous commitments

with enigma and secret network what is the commitment here u considering that >> I mean I stopped working on secret network in 2021 I do this full-time now >> okay >> that is time uh thank you so much John please give a big round of applause >> sorry I didn't give any time. >> No, no, no. >> All right, we are ready for our next team. And for this one, I'd like to bring on Glider. And to tell you all about Glider, please welcome Brian. >> Well, hi there. I'm I'm Brian Huang. I'm one of the co-founders of Glider.

>> Uh, Glider automates any investment strategy. >> We're a team of DeFi veterans who have been in crypto for at least a decade plus. I'm a former high frequency trader from XTX Markets. I also led the trading product at Anchorage. My co-founder, John, built and led all of matcha.xyz in the early days of Zerox and we're backed by some of the biggest consumer brands in all of crypto. Um, fundamentally we believe that the way people investing is changing. And what do we mean by this? So, in the 70s we had index funds that became ETFs. We saw direct indexing and zero commission in the 2010s. And now we truly believe that we can build personalized portfolios around anybody's goals, positions or themes that they

want to invest in. In parallel to this, we have the replatforming of every single asset possible. And so RWAs are on an absolute explosion from basically zero in 2023 to over 30 billion dollars of assets now. And so when we think about the solutions for what exists and how you want to manage these assets, unfortunately, we're left with this. And this is not where anybody is going to manage their corporate bond equity portfolio for real investors in the long term. And so not only that, that's just the first step, right? Once you get a wallet, you have to fund that wallet. You have to figure out where the app is. You have to get to an app. You have to manage your

position. I think we can all agree that no one wants to use these products. That's not the way that people generally invest. And so, let me introduce you to Glider. Glider is the calm, intelligent way to invest on chain. And so, with Glider, you can do things like, I want to build a portfolio around healthcare stocks or airline stocks or defense stocks or corporate bonds, any chain or rather any asset that is on any chain. And you can do this all without the complexities of DeFi. So no gas, no signing, no bridging. It's all been abstracted away. And so our goal here is to build the execution engine and orchestrate the movement of all of these assets from point A to point B without you having to figure out what that means. I'll give you an example. If a particular stock

reaches a limit price of 200 bucks and you want to sell that into corporate bonds that are then earning 7% yield, that happens automatically through Glider. You don't need to do a million transfers between Coinbase to Robin Hood to your Bank of America accounts. Um, we launched about 6 months ago at the end of November. We have over 100,000 users. And uh, the most important thing that we've been launching recently is invest on autopilot. And you might be like, well, don't we already have that? And the answer is no. It's only available to people in the US. And so all of these portfolios that we've built on Glider, most of our audience is actually outside of the US and this is the only way that they can access things like the Nancy

Pelosi tracker or build me a portfolio around biotech US stocks. Um, not only that, these portfolios can live in any interface. So I can take a Nancy Pelosi portfolio and I can offer it into Phantom, OKX wallet, some of these uh wallets and neo banks that we've been working with behind the scenes. Um, and with that, comment glide with us. Thank you. >> Amazing. We're gonna go kick off our Q&A. So, please take it away, judges. >> Hey, Brian, thanks for the presentation.

Um, maybe talk us through how you got your initial users and how you've grown. It sounds like you have a 100,000 users. Is that what you said? >> Yeah. Um, so our first fund raise really targeted consumer brands for a particular reason. um we had everybody tweet about it and I think that's what grew the initial interest into the product. But right now what we're seeing is incredible demand out of Asia. Like as of two three weeks ago, China banned the ability to invest in US public equities. And because of that, they are now looking for alternative methods to do that. In particular, onchain stocks provides a lot of that. Now, you could argue that onchain stocks investing is quite commoditized, but there is nowhere that you can actually build a

diversified direct index portfolio besides Glider. There's not a single product. >> What are some of the most popular products like besides the one you mentioned from copying trades from >> Yeah. So, all this is fairly new. Like we've been around for about a year and a half. We launched in November. In November, tokenized equities were relatively small. And now we are seeing people do things like I want to hold maybe the core four which might look like spy, Apple, Nvidia and Tesla. Um we're seeing a lot of people build things around preIPO stocks as well. So I want to hold anthropic uh OpenAI and SpaceX. I mean anything that is on chain can be built into a diversified portfolio that people generally want to invest in.

So basically it's a non-custodial solution. So that basically still means that I have to delegate execution rights. Do you what would be like the worst case scenario uh for my wallet and what could happen to it? >> Yeah. So you don't uh you don't run glider on your main wallet. It is a non-custodial wallet that you can think of as like a sub account to your wallet. You have to deposit into it. And so those funds are allowed to do certain things like swap, like lend, and that's codified into the session key that's on the non-custodial wallet that you sign.

Um, you know, if Glider ever disappeared, because these funds are non-custodial, you can always claim your funds back. You can always revoke session keys. Like there is no sort of counterparty risk to Glider at all. >> And on on the demand side, who do you think which side of the business is pulling numbers? Is it consumer side or API side? Yeah. So everything that we've done so far is direct to consumer. Um B2B we think is really interesting because that's a million different distribution channels and these things can be spun up in an afternoon. So again, if you want to offer the Nancy Pelosi tracker on your neo bank, you can do that and it's actually far more user retentive. It generates volume and it generates revenue as opposed to swapping. Nobody wants to swap on a neo bank because there's already 15 million

places that can swap. These are like ETFs. people deposits, they forget about it. It's a good investment long term, especially with stocks, and it just generates revenue. >> Thank you. >> Um, so you know, you've been around since November. What are sort of like user retention numbers? How much churn do you have? How do you think about cohorts that are sticky versus cohorts that are turnurning? >> Yeah, so our like 120day user retention is about 86%. Um, again, this is all set and forget. So you deposit into these diversified portfolios. The money sticks around. We ran a campaign with base um two months ago and 82% of the funds stayed within the portfolio after the campaign had even ended. So three weeks

after the campaign had ended. Um and I think this makes sense, right? This is the way that people invest into ETFs. They park money and they leave it there and it continues to grow with them. >> And I guess a final thing is how do you see the long-term competitive landscape? How do you view like margin compression between these apps versus sort of like order flow based stickiness like you have with Robin Hood versus >> So we control the order flow which is great. Um you know I think about the incumbents in the space like Robin Hood and Coinbase they're bounded by geographical constraints. So I would argue that that's not incredibly competitive and in fact we sell into them B2B and then globally um I don't see an offering quite like this yet. So it seems to be the only solution for a

lot of this like direct index diversifying. You could think of this as a layer higher even too where people like Bitwise, Fidelity, Wisdom Tree want to actually offer ETFs on chain and use Gliders infra for that. All right, next up we're going to bring on Zyi. So, please welcome Guthia from Zifi.

>> Does it work? Oh, yeah. Perfect. Hey everyone, I'm Goautier, founder at Zyi. Uh we've been building products in this space since 2021 in DeFi and account abstraction. Um and today we're going to explain you why we've built personal yield agents embedded into users wallets. So let's look at how DeFi is constructed today. You have the protocols, you have the risk curators, and then you have the users. And today users have to pick the markets, monitor positions, rebalance their funds to make sure that they're always generating the best APY. and the user experience is very poor. And how do you make that simple? How do you make that accessible for everyone to access? Well, you need an agent that can abstract all that

complexity. And this is what we've built. Personal agents that can manage wealth on chain for you. I think vaults were great in the early days because it was abstracting complexity, but it was one sizefits-all type of strategies. Here with agents, users can customize their risk profile. They can customize how they want their funds to be allocated. Everything is automated and things can be monitored on behalf of users. So this results by us improving the native yields of users trying to generate yield in DeFi by over 48%. So as I'm speaking right now, our agents are are continuously monitoring users position and rebalancing them and

they're also making sure that they're always safe. So, we've seen the recent DeFi incidents with all the hacks. Uh, and some of our users were saved thanks to to our agents that are rebalancing continuously users funds. So, how does this look like in real life? Uh, we've been live for over 12 months, we've rebalanced over $2.4 billion of funds. And as I'm speaking right now, we have over 10,000 wallets which are being monitored and which are being rebalanced actively. We have around 6,000 monthly active users, around 7 million asset under management and taking a 10% performance.

So the market is huge right now. The yield agents um vertical we have around five to six projects which um are competing with each other and we represent around 25% of that market. Um and I like to compare what we're doing to managed accounts but onchain. So there's huge room for that to grow. uh 14 trillion markets as RWAS and equities is coming on chain. Uh we have around 8 million that is committed and incoming uh for people to to use Zyi and integration with different wallets and creators. So we've just started with yields but we're going to go beyond that and provide managed accounts on chains for everyone. So DeFi is becoming a aentic. So if you want to help us scale

and grow uh you can shoot a message and and also try the app if you wish. So thank you for listening. >> So Yan um so agents historically are non-deterministic. Um yes how do you how do you manage that? Yeah. So the execution is fully non uh it's fully deterministic. But then we use the agents to find new opportunities to monitor the markets, monitor like the collaterals inside of different vaults, collaterals of those collaterals, give specific ratings um and also improve the algorithm itself that helps rebalancing the users's funds.

>> And um in terms of uh agents losing somebody's funds, where the liability lies? Is it on on you on >> I I didn't understand the the question. >> Um so if the agent's losing money if there's any breach where the liability lies. >> Um yeah this is a bit of a of a gray zone. We didn't have any incidents so far because you know the it's an algorithm which is doing the execution. So it's it's just always doing what it's supposed to do. So uh we didn't we didn't have any issues until now. So, but it would be a bit of a gray zone through the fact that it's the agents that's could be liable. Yes.

>> Yeah. And uh by the way, just uh really like that your auditor is Sakura, which is our portfolio company. Uh appreciate appreciate that. >> Oh yeah, they done a great work. >> Uh my question is you said you guys own 25% of the market share for the Gent yields. Can you talk about how you compare with the other competitors? I don't know much about them. >> Yeah. So there there's I would say five six other projects which are pretty much doing the same thing. I think we've been leading right now the the vertical with u you know we have a huge community people talk about us on Twitter we provide the highest yields. Uh we've been able to really react quickly with all the recent uh DeFi incidents with

like USR DPEG a incidents. Um so I think we from at least a public perspective we're we're leading the the vertical right now. Um and also in terms of numbers we're yeah leading as well. >> So say like an institution comes in obviously if there's a institutional partner that comes in wants to deposit 100 million. >> Yes. This is this is the next steps for us. We have like 8 million which is coming in and we're starting to to talk with the different institutions >> and um trying to figure out how we can improve all the risk monitoring systems, security, uh everything just to bring more trust around what we're doing and grow the TVL to to be able to acquire

these these types of players. >> Maybe tell me about where the 8 million is from. It's on chain wells, web three wells that >> Okay. But not like I'm talking about like institutions from you know traditional >> not yet. That would be the next step. Yeah. >> Okay. >> Yeah. Yeah. >> Um what are the most popular you know you you started by talking about how you know the strategy space is kind of boring over the last year. What are the most popular non-trivial strategies that your users use? like not just simple yield farming like yeah talk talk about the strategy space because I think that was not totally clear to me.

>> Yeah. So right now what the agent is doing is just rebalancing between uh different lending markets and vaults. Um so it's it's pretty basic in terms of the strategy itself but we've seen a lot of demand for like looping um even like LPing like there's so much to build on top of this right and and we we started with lending because uh it was the use case that made that made the most sense for us in the early days as you have to monitor all these different pools collaterals risk creators um and yeah this is just to start with lending markets but you could go way beyond and and and provide automation for like different type of strategies like delta neutral lping and and all

that's in in DeFi right now. >> Yeah. Uh and what's your roadmap for those strategies like what's the next set of strategies? >> So we have a lending agent. The next thing is a borrowing agent. So so to find the best borrowing rates and and be able to rebalance those borrowing position through flash loans to then lend out at the best rates. And then eventually we could introduce looping as well. That's on the next steps. Yeah. >> Thanks. >> Right. We are at time. Please give Gothy a big round of applause.

>> Thank you everybody. >> All right. Next up, we're going to bring on ASI. And to tell you all about ASI, please welcome Aniset. >> Hey everybody. We're ASKI and we built Box, the most powerful and affordable sandbox for agents. You see this street? I believe in two years, everyone on it will have agents working for them 24/7 in the background.

But for this to happen, we need to give every single agent dirt chip and powerful computers. one trillion dollars in 2025. Laptops, desktops, and smartphones sold for that grand total worldwide. For machines that humans use to do work in today, everything is computer. Whatever you do, I believe for agents, it would be the exact same, but at such a larger scale because agents never sleep and you can replicate them as many times as you want.

If humans need laptops, AI agents need computers. And those computers we call sandboxes. Today, sandbox providers in the world sell for more than $100 million per year. And I believe this is such a tiny fraction of what this market can grow out to be because in the end it will be these consumers that will be using consumer agents that use computers to do work for them. But according to builders and companies making consumer agents today reaching out to me every day, you can't afford more than $1.5 per computer per agent on consumer

products. Otherwise, you just can't make the pricing work. That's where we come in. We're a team of two young technical mavericks from Europe. And we spent the last nine months building the most powerful cloud coding agents out chipping Devon and cursor on capabilities silently. And from that we've extracted box the best compute primitive for agents to do work in. It comes as a CLI an API and SDKs for developers of products to give um computers to their agents.

We have all the state-of-the-art capabilities that everyone else in the market has. uh desktop use, virtualization, dev tooling, forking, whatever you want. But we give it at up to 25 times less expensive than all our competition in the United States. And on our benchmarks, we rank first on how easy it is for agents to build complex software with. With us, it takes less steps and there is almost no errors as well as being very cheap. Really boxes are paradise for agents and we're the smallest leanest team in the world to have ever achieved to build such a product. We're funded by Bay Area

investors and three weeks ago we launched on Twitter. It went viral in our niche on Twitter of AI agents builder and we now have customers that are renting thousands of computers uh for from our platform and two of them were so convinced that they've even angel invested in us. So if you're building agents or investing in agents come talk to me. Thank you very much. [applause] >> Amazing. Let's go into Q&A. Um thanks for for uh that. Uh so you know you talked about the uh 25x cheaper um kind of performance. Where's that

coming from? Is it coming from token routing? Is it coming from like sort of you know how you're managing tool calls? Walk me through the optimizations. >> So we don't sell tokens. All right. It's bring your own agent. Um we sell the compute that the agents does the work on. Right. Your agents in the computer. We build you by the second for how long the computer is running. And why we're so much cheaper is that as I said, we're the smallest Linux team and we're from Europe. Um, and we really have almost zero operational costs. But on top of that, we are the first to build it on top of HNER, which is the least expensive cloud in the world uh in Germany and uh we have agreements with them to use a lot of their compute.

Um and then a second question is you know the agent landscape has been getting more competitive right obviously you know between Hermes and and others um >> how do you place yourself amongst the landscape of people building you know >> agent platforms like >> openclaw Hermes etc versus people sort of building um you know pure compute sales you know you know it seems like this is sort of somewhere in the middle right it's like you're I'm not getting just a raw box, but I'm also not just getting a full agent. So, yeah, how do you how do you see sort of the competitive landscape? Because I could imagine, >> yeah, >> people who are making agents would move up and start selling compute. You already see a little bit of this.

>> So, yeah, walk me through how I mean, historically, sandboxes um their biggest sales were for reinforcement learning. So, they would be selling to big labs. Uh but now we see this new wave of uh agents becoming intelligent enough to use advanced tools that require an entire computer to to work. Um and so this is more more of an emerging market even inside the sandbox market. Um in this we're amongst the founders that have the most Twitter following and everything there happens on Twitter. Uh and we're very good at launching at getting inbound there. Um yeah I think I think we can do we can do well in that.

>> Thank you. >> Yeah. So my question would be where's the crypto angle here? Um >> yeah there there is none. >> Thank you. So why ATH cons then? my co-founder ETH global uh we've we've met together four years ago doing an exchange in near uh and one of our angel investors even was a exar member uh so we're kind of close to crypto and honestly if you build agents in crypto you probably need languages like rust maybe you need chrome to go and gather information and we're one of the sandbox providers that provide that >> yeah and my second question u what share of agent output actually ships to production to production repos any

>> as I said we don't provide agents we provide the compute for them to do work in >> so this is to the discretion of our users >> thank you >> uh last question for me it sounds like you've gotten your first users and customers from Twitter how are you looking to scale next >> um yeah so basically the way you scale this is that you find the couple of customers who are building platforms that need a lot of agents we're currently for instance in talk with startup in Germany that wants to and 3,000 agents at the same time all the time. This will bring us to 1 million AR R. So then you find a couple of these companies and you scale like that and you go further and further and further in demand. But then the long tail of people who discover you bottom of market

is big. But because you want to insert yourself, you want to impregnate these kind of companies that use a lot of agents. >> Amazing. Thank you. Please give a big round of applause. >> Well, we're going to get ready for our next team. And for this one, I'd like to bring on paper. And to tell you all about paper, please welcome Chris. Hi, I'm Chris Sove, co-founder of paper and former practicing attorney. Paper is the agreement layer of web 2 and web 3, allowing any legal agreement to automatically execute its terms.

For 4,000 years plus, human beings have put their agreements in writing on tablets with uniform on paper with pen and now PDFs with eign. But nothing has changed. I can print a PDF, sign it with a pen. It works the same as any other agreement over those four millennia. The problem is that the agreement stops after the signature. And when there's a problem, it often costs more to enforce the agreement than it does to walk away, especially for small businesses and in international transactions.

Until now, the legal meaning of of uh legal contracts and the automatic execution of smart contracts have not come together. Paper brings those two worlds together, meaning and execution. We've already implemented in the real world with ETH Detroit. ETH Detroit's pizza party last month used paper to serve their liability waivers. Dozens of liability waivers were signed on site flawlessly. Everyone scanned a QR code, read the terms, agreed and signed on chain. But the bigger unlock is payment.

For instance, a software contractor and client can enter into a milestone agreement. The con the uh contractor enters the terms, the legal terms, they create the agreement and it goes on chain. The other the other uh the um client proceeds and they have a legal agreement signed on the blockchain. Now we hope that the agreement goes as planned. We have a couple of milestones. The milestone is delivered and the uh client agrees and funds are released automatically. But if there is a problem, they can initiate arbitration online so they have certainty. They're not getting courts or lawyers involved.

Now is the time for paper. We have an audit log that shows every step of the process. And um users sign in with with email. They avoid wallets and gas, but cryptonative users can bring their wallets and their crypto to these agreements. Paper is the agreement layer of the internet. Come visit us at our booth or visit papap. Put your agreements on paper. [applause] >> Hi, thank you for this. Um I'm curious what happens after you click on initiate

arbitration and what happens in that flow. >> So paper is modular and extensible. So right now if you hit hit the button we have just three options. There's single arbitrator, panel arbitrator and then the third one which gives more time to introduce evidence etc. It's really an example that's not live in production right now. Um because we need competent legal counsel. A lawyer who has himself for a client has a has a fool for a client. So I I am not legal counsel for us. So we need to work through that with with legal counsel. But it's infinitely extensible. the parties agree and have certainty about what style of arbitration, mediation, etc. is used.

>> Got it. Got it. And how do you want to introduce paper to customers and clients? Is it going to be a platform where like docuign where folks are going gathering and signing documents? How do you think about that? So you you picked it up and this is a new development for us because this can go anywhere and what we have heard over and over again at this conference at European blockchain convention other conventions is that that participants in conventions like this want a replacement for docysine and so we are starting though we can do much more we're going to start with a docuign replacement paper sign and go from there but paper sign is one clause we call it or one agreement on paper that's one use case the docu sign replacement. So that

one use case then we extend from there to other use cases. >> What sorry last question um what's the most complaint like the main complaint about docuign that you can replace and really be the better alternative? >> Yeah sure. So, we have a very very thorough audit trail and um and it's a permanent auto trail audit trail that is not controlled by a centralized authority like Adobe or it it's on the blockchain. Anybody can view it. A court can view it.

>> So, I I had a question about um the translation layer. So it sounded like you have something where I can take a legal doc in text, you know, put it through a an AI model, generate some representation in your kind of mapping to your framework. >> Yeah. >> Um how how do you audit the correctness of that mapping and like how do you how do you do things like okay maybe I have a a a legal contract it has an escrow quantity and then it says okay under these conditions the escrow can be released. How do you validate that the economic flows mapped in the smart contract actually map what the the >> so we are we are providing the rails that these agreements run on. Right now

we create the cars that run on the rails but eventually others will create those agreements. >> So they're like your initial one are some templates. Okay. Yeah. Yeah. That's the and they map. I can tell you how they map. So it it most here may know this but when an attorney creates a document they copy and paste from another word document clauses it's the same with paper being a former attorney um we have clauses but that make up our agreements >> and there's four parts to each clause that are possible they don't always exist onchain offchain clause which does offchain processing and UI and so those come together there's a manifest

that ties those those four pieces together and I think I missed one of them but those four pieces oh the legal agreement side of it so those four pieces come together in that manifest and we don't necessarily validate it now I've considered using a to validate but we do not necessarily validate it yeah >> and and for the final question I was for like forking you know I think of legal documents the copy pasting is like I have some some forkability and also replacement yes how do you do deal with like replacing you know a document that's a successor of a previous one.

>> Yeah. So when you say successor a it replaces it um >> succeeded. Yeah. Yeah. Um there is a link between agreements that already exists. You can link to a prior agreement and uh it's possible to trail back through the chain of agreements. Incidentally also offered offers and counter offers can be tracked and you can go through that chain of counters offers and offers which is another innovation over docuign. >> Yeah. >> Yeah. I have two quick ones. Hopefully I have enough time. So correct me if I'm wrong. So in general for those contracts to have uh enforcement the companies will have to put money into the escros.

>> Mhm. That's correct. >> Okay. And then is there any onchain component to it? Uh do you do agreements bound to vesting or signatures or >> Yeah. So absolutely. So anything that self-executes will be part of a smart contract that we call an agreement. It's signed on the blockchain. The milestone agreement is on the blockchain. We expect that our customers will mostly use US dollar coin. We want it to be denominated in dollars, but they can use crypto uh balances to denominate their contracts.

>> Do you believe that is a capital effective for companies to put money into the escrow all the time for it to be kind of bounding? >> Yeah. So, that's definitely a consideration. Um, this is not my area of expertise, but my my belief is that those funds can be invested essentially while they're held in escrow and they get to take the the proceeds after that milestone is done, but that's beyond my expertise. >> Yeah, thank you very much. I believe >> we're at time. Please give him a big round of applause.

>> Thank you. >> All right, we are gonna just jump right into the next team for this one. Let's bring on Ascent. And to tell you all about Ascent, it's Killian. Please welcome him on stage. Thank you very much. My name is Kian. I'm the co-founder of Ascent. And our goal is to bring market making yield to every dollar. Uh so every tradable asset needs a market. And all of these markets need a market maker. And right now in traditional finance, only a handful of firms uh basically dominate this market that is highly competitive and they take home all the profit behind closed doors.

So when DeFi brought uh market making onchain in automated market making we let anyone play the same game but we didn't give them the proper tools and we face slightly different problems coming on chain. So for example liquidity providers can't set their spread so forced to pick on static pricing just leaving profit on the table. Next, we open the door to MEV bots to exploit the swappers and the LPS. And we also have a lot of passive capital that just goes elsewhere in DeFi because the risk of impermanent loss is too high in market making and uh they seek yield in other places. As you can see, almost half of all LPs actually lose money from liquidity providing on chain.

So, this is what Ascent fixes. Uh we've introduced a pricing model that reads incoming swaps and we classify them based on if they're toxic or if they're healthy flow and we price them accordingly. And then once we've uh once we've priced it, we then auction off the arbitrage to the bots in order to rebalance the pool and properly compensate our LPS or those fees. And this is what it looks like in one of our pools. The top the top example here is a fixed fee tier pool where the arbitrageer is taking most of the profit and we're able to capture almost 80% back from the the MEV arbitrage and distribute that to our LPs in our pools.

So the pool performance kind of speaks for itself. Uh we've already done over $1 million in uh trading volume since we launched these pools in May. And as you can see our stable coin yield is is uh like through the roof. And uh but it's not just pools either. We're hoping to launch some vaults uh that will be a one-click deposit in order for uh passive investors or um in order for capital to come in and be able to have access to some of these yields without all of the market making risk involved. So, a little bit about our team. My co-founder Kevin has a background in investment banking. He's been building structured financial products for the best part of a decade. and myself uh come from a technical background as an engineer and we've been running

successful DeFi strategies as LPS uh for 2 years now before we started ascent and uh won the unis swap hook incubator program and uh so we have all this liquidity coming on chain and we want to make sure that the infrastructure is is built right so that we can make efficient markets and we don't make the same mistake leaving market making up to just the same handful of firms who dominate the space right now in traditional finance Um, so if you're an LP, please feel free to check out our platform. You can explore our pools. Uh, if you're a vault curator, run your strategies on our pools. Um, if you're a token issuer, we, you know, we'd love to chat with you if, uh, if, uh, you're looking to, we can launch you a custom pool or if you just

have payers that that you're interested in, uh, please reach out. So, thank you very much. >> Yeah, thank you very much for the presentation, man. So what signal actually moves the fee volatility? Is it like a flow toxicity in Oracle? And how do you avoid being gamed by searchers and good? >> Yeah. So we we can never fully avoid being gamed but we use we don't use any oracles. So we price the swap before it comes in. So we use mostly the the price of the swap and how far it will move the price and then the volume and the frequency that it comes in with this time decay. Yeah. And just in general in business model, who pays you uh is it a

fee card or on pools or licensing or any any other model? >> We we take a protocol fee from the LP fee. So we're never charging more on the swap. We just take a small cut of liquidity provider fees. >> Yeah. Thank you. >> Thanks for the presentation. Um so I guess my first question is more a question about um you know scalability. like I think the initial assets you have are sort of majors. Um and so I'm just sort of curious like in your mind what is the ideal composition of majors versus altcoins and and sort of like how do you view sort of the the tradeoffs there because obviously I think like you know these

toxicity signals will be very different for low low liquidity tokens. So yeah walk me through how you're thinking about scaling. >> Yeah. So I I think a lot of these altcoins face even more volatility and a lot of LPs are facing even greater impermanent loss especially if the assets are correlated or uncorrelated. So I think um these pools are especially ripe for a custom logic that will compensate LPS properly for the risk they take going into riskier assets and riskier pools >> and and walk me through the architecture. Is this you know you you mentioned the auction so there's sort of some asynchronous looking component but then an execution [clears throat] component. like what's sort of the full flow of infrastructure and >> Yeah.

>> Yeah. So, we're pricing the we're pricing an incoming swap and then uh preceding that swap is when we auction down the next swap. So, the arbitrage or the the gap between our pool and let's say another pool and um whenever the bid is looks uh appetizing enough for an MEV bot or an arbitrageer to come in, they'll take that. And do you worry about you know say during like a very large downswing uh when solvers are sort of down that you'll be sort of your LPs will be sort of taking implicit toxicity because because no one is taking the other side like how do you think about the like lack of participation impacting performance like you know say there's like a October 10th style event again right where and where basically no

solver is uh fulfilling any orders like how do you how do you think about >> preparing your LPs for that? So we're we're a v4 pool on unis swap. So we're built on top of unis swap's architecture. So um but we integrate with other routters. Um I'm not sure if I can properly answer the insolveny question there. >> Yeah. No just more like you know there were certainly a lot of RFQ style systems where or like sort of interlace systems where on October 10th LPS took you know they had amazing returns and then obviously it lost everything.

So we would hope that our logic would kick in properly and and scale the fees up to even the maximum. >> Yeah. Cool. >> All right. Thank you so much, Killian. Please give him a round of applause. >> Thank you. >> Next up, we're gonna bring on Cambria. And to tell you all about Cambria, please welcome William. >> Hello. Yeah, I'm Will and this is Cambria. We're at Crypto Game Studio.

So, if you don't know, gaming is actually pretty big. Um, it's actually bigger than film and music combined. And so, we, you know, started making sort of small mini games. Um, one of the big things we've proved is, you know, risking money on chain and then tying it to in-game. Um, it's a very simple miniame, but you know, just two contracts. We're allowed to do peer-to-peer duels, and yeah, we got a lot of unique players, a lot of risk, ETH, and then we've expanded to Gold Rush, which is sort of like a big bigger MMO experience, really trying to nail like,

can we make a fun game that's, you know, massively multiplayer, pretty hardcore to be honest. And yeah, we did we did great. Um yeah, Cointelegraph shout out. Yeah, Jihas, uh the founder of Ronin and Axi actually gave us a great shout out on uh our season. And yeah, we got a uh a lot of good feedback that you know you crypto people which are our main sort of initial audience, you know, they can talk to their old Runescape friends, bring them in, we can sort of break out of the bubble.

And so yeah, we have a upcoming product just came out our launch video. And yeah, it's been going pretty well.

Um, it's been live like two, three weeks, you know, good retention, good volume, and yeah, we're just, you know, trying to build it out, trying to make web two onboarding a lot easier. We still think crypto gaming is like a great onboard for crypto as a wider sense. Um, yeah, people just love playing games right? And yeah, we're also been focusing a lot on like worldwide expansion. Another great thing about crypto being worldwide. Um, and yeah, so we're trying to carve out our own sort of slice here and expand in this expanding market.

So yeah, this is Cambria. Yeah, our sort of final goal was to, you know, get to like a Wowsized or Runescape sized full MMO experience. Thank you. >> How long have you been building Camra? >> Little over three years. >> So, this game that you just demoed was a three year kind of >> um so the most recent like the video. >> Yeah. >> Uh we started working on that probably like six, eight months ago. We sort of have like a common engine and uh yeah it's like an amalgamation but like yeah like dual arena has been wide for like two years. We sort of took it offline.

We sort of have a carousel. We'll update it, integrate it. Eventually it'll be sort of one big common world >> and the whole game runs on the abstract blockchain. >> Yeah, we have it on abstract. We've been on like Ronin in the past. Um yeah, >> I mean it's still a game. Uh we have like a you know servers and all of that. Um >> but yeah you know >> so which >> anything financial basically anything financial I think. >> Yeah for instance like in gold rush to join you had to buy a charter to play or sponsor someone else's charter. Um so that was on onchain item you know it makes an event our game engine picks it up. Um and yeah all the price cool logic and everything at the end was on onchain

and distributed onchain. And yeah, it sort of like backs the economy, right? Like the in-game economy is backed by this price pool. So you get some interesting effects. >> And do you have a token incentive for games right now? Do you incentivize it right now? >> Yeah, a little bit. We have sort of like a points program. Um, yeah, token coming soon hopefully. >> And then when the token incentives cools off, what going to happen next? How do you incentivize people to actually come and play? Yeah, I mean I think the game just has to be fun, right? You know, plenty of games. Yeah, it's sort of like a traditional game, you know.

>> Um, my second one, a pretty general one. So, usually real money attracts cheaters and what farms and how do you how do you deal with like the anti-che approach in that sense? >> Yeah. Yeah, we have to do a lot of monitoring. Um, yeah. I mean, it actually shows up the numbers pretty easily. um just sort of like added distribution. So yeah, we just are monitoring that sort of all the time. Uh that's sort of one of the reasons we did like limited seasons for gold rush is like if it's a week long, you know, it's not a critical it's not as critical of an economic issue. We can reset at the end of the season. Um yeah, definitely something we're just working on live ops, you know, getting our tooling up, a lot of behind the scenes

stuff. >> How many I missed. How many users do you have right now? How many gamers, players? Uh like right now online we have like over a hundred. You can actually yeah sign up go to the link play. Um yeah. So you're talking right now all time >> in general live now and all time as well. >> Yeah I think live now is like a little bit over 100. Maybe it's different. I don't know. I haven't checked. Um >> but yeah we've hidden like I think our peak during gold rush was four and a half thousand concurrents. So yeah everybody you know playing the same game at the same time for a week.

And on on revenue side uh sorry >> can you on the revenue side can you share the numbers like approximate ones how much do you generate per month per week per day? >> Yeah it depends on the product um we sort of taken a different approach depending on what game um like for DS we just take like a 2% fee sort of on all the volume. Um for gold rush similar just percent of the price pool. Uh, and yeah, dungeons sort of the same way and the payout. So, yeah, we're just doing all like volume based. Um, we think that's sort of the most aligned with our player base.

>> Okay, thank you. >> Um, I just had one question which is what is the risk to earn mechanic? Like I'm locking up USDC and then I'm allowed like Yeah, walk me through the mechanics of of that. >> Yeah, so I I mean it's really like PvP at the end of the day. Um, we think that's sort of the most sustainable, you know, for duels. Yeah. You're depositing ETH. Your counterpart is depositing ETH. Yeah. You're fighting over it. One person wins. You're risking capital.

>> Got it. Got it. Thanks. >> Yeah. >> All right. We are at time. Please give him a round of applause. Next up, we're going to get Frank and Coin on stage to tell you all about Frank and Coin. Please welcome Johannes. Very good morning. Hi, my name is Yhannes. I am the managing director of the Francoon Association. Frank is the largest Swiss Frank stable coin and today I'm going to talk about the Franken pool shares which is the governance token behind this all and this lets you own a stake in decentralized Swiss Frank money.

Now 99.8% of all stable coins are in dollar. Now that might sound really good because you have a lot of liquidity. You can trade, you can get go in and out and send it around the world. But to hold those assets is actually not really that good. In fact, the dollar has lost more than 80% of its value against the Swiss Frank in the last 50 years. Now that's not a secret. In Trati, the Swiss Frank is actually playing this outsized role and many million institutions are holding it as this store of value. With Frankencoin, we're bringing this onchain.

Frankcoin is life and scaling. In fact, it's the seventh largest non-stable coin at the moment with about 32 million Swiss Franks in supply backed by more than $70 million in assets. Frank is dominating the Swiss Frank stable coin market, but the opportunity is much much bigger in a stake environment. The opportunity is quite clear that we need something else as store of value. Now, how does it scale? Frankin is built as a CDP but its oracle design is really really well suited for RWAs and fits perfectly in is quite unique market the Swiss Frank moves in. Francoin is integrated in many places already. You can pay natively in the third largest

supermarket chain in Switzerland Spar with it. You can use a Visa card with noises pay. It's integrated all across DeFi and it's fully compliant in most jurisdiction in European Union. It even has a ma white paper so it can be listed on any exchange without any restriction. Now the Franken pool share the Franken Poolshare is as I say the governance token which allows you to own part of the system. You can participate in the governance and you get in on the upside. It's currently valued at $16.3 million and last year alone it made almost $800,000 in uh dollars in revenue with a monthly growth rate of the underlying of almost 26%. To sum up, stable coins are

still mainly in dollars. The Swiss Frank is actually much better store of value. The Frank coin is here. It's live and the potential is massive. With FPS, you can get in and you can part be part of the upside. come and own a stake in decentralized Swiss Frank money. Thank you so much. [applause] >> Um, thanks for that. Uh, so quick question. If I remember correctly, like Swiss Frank native yield from government bonds is like sub 50 bips, sub 25 bits. >> So you're viewing this as a pure payment stable coin, not sort of like something where there's real intrinsic yield. I it sort of seems hard for me to imagine

compared to say Euro C right where there is some native yield from Euro government bonds and then hopefully offchain activity how how do you view like the yield kind of world >> growing there there is yield on it there is also three and a half% yield on right now which is dynamic but it's three and a half% yield on the Franken coin now you're completely right that yield does not come from government bonds because they're actually at zero or even man negative uh rates so it is a CDP so the yield comes comes from RWAS. It comes from bonds, uh, real estate and so on. It's structured a bit like a a bank. You could could argue, >> but but your carry cost for I'm assuming that Swiss den like CHF denominated RWA's supply is like significantly

lower. So you're going to have some carry cost for having to pay for dollar RWA. Okay. So your your next step is CHF denominated RWAS, >> correct? >> We have um so Switzerland is just a a side project. Switzerland has really really good regulation for that. So there's the DT law in Switzerland which actually allows you to natively issue for example equity as a token. It's not tokenization. You literally make a capital increase and issue it as a token. So that makes this whole setup very cheap, very efficient and you can scale it quite nicely. Currently have three uh larger Swissmemes as collaterals in there. But uh we're we're working on literally dozens of projects to bring larger assets in here and to scale this up. And uh final

question is um are there sort of neo banks wallets that are like natively using this locally or or like is there a pipeline of that? >> Yes. Yes, there is. So there it's it's fully integrated in seal which is coming off ecosystem that's used very very actively. Um you also have free on and offramp. So there is kind of a a zero cost entrance into the local banking system out of the local banking system. Um, and there is more neo banks actually being developed coming. Um, and there's integrations coming up. >> Oh, actually, sorry before you, what's the on-ramp cost? Median on-ramp cost.

>> Zero. >> Okay. >> But, but at all scales, if I wanted to do 100 million CHF, >> um, that's typically about 0.3% at the moment. >> Cool. >> We're working to lower that. >> So, I suppose it's an Oracle free design that you have. >> Excuse me. Oracle free design. >> Yes, it's an Oracle free design >> and I I believe it's the biggest differentiator but at the same time puts you in the in the big risk. So how do you resolve that? >> I would say the opposite really. It's like one of the big upsides on the risk side because oracles are one of the biggest risks together with um correlation traditional CDPs have and by removing this CDP sorry this oracle risk um you remove that risk but you also add

in um a wider universe because what oracles imply is that you need very liquid immediate markets and deep pockets because you're usually liquidating into very unfavorable markets when you have to liquidate and frank kind of puts it on its by um being slow instead of fast. So it's built to survive slow liquidations. That's intentional. It can absorb those losses, but it will bottom line have less liquidations. And actually um if you look at what happened the last three years when it's been live, they've been about 50 liquidations. Um all of them have been super profitable for the system down the line because the system works and the market behaves a way in a way that those oracles actually are super efficient. Sorry, those

liquidations. Oracles. >> Thank you. Yeah. And just Okay, go ahead. Now, in terms of the Swiss Frank, um, so the the demand for USD is like way higher. What do you believe like the top ceiling for the demand on on Z C HF? >> Sorry, how the demand looks like? >> Yeah. Yeah, >> really big. >> It's really big. It's uh actually the demand side is not the the challenge for us. It's as uh Aron very very nicely said it's we need to keep up on the supply side.

>> Okay. >> We could grow a lot quicker if the RWA's market would bigger and I hope you know everyone is building here built quicker like [laughter] we we have demand we have uh massive demand on on the just people wanting to hold it want and the polit it's a political play in the end right people wanting to hold an alternative to the US dollar full stop. >> Yeah. Thank you. >> Amazing. Thank you very much so much. Next up, we're going to bring on Yield Seeker. And to tell you all about Yield Seeker, please welcome Christian. >> Hey.

>> Hi. So, I'm Christian. I'm here to tell you about Yield Seeker. So, our platform allows you to grow your wealth on autopilot with DeFi agents. Our thesis is quite simple. We think DeFi is better with agents. Uh, and let me show you why. So, we all know DeFi is incredibly powerful onchain systems. uh they offer direct direct permissionless open financial operations and we can use them right now and more and more of trifi is moving to DeFi. Billions in RWAS are coming on chain and these will all be accessible with programmable primitives but executing really well is really really hard. Uh a simple example is earning yield on stable coins. You can do an easy task and earn 3% but to earn more it becomes very timeconuming and very complex very quickly. Here's why.

like everything is fragmented. So you have to check multiple platforms. You have to check them daily because the rates will change quite constantly and sometimes quite dramatically. People give tokens and uh rewards and random tokens you may never have heard of. Uh fees are hidden in drop downs like it becomes so complex so quick and this is the simple case for really complex things like complex trading trading strategies. It really is really far hard for use users to use like it's far away from their abilities. Uh and the problem is only growing because as more and more real world assets come on chain, we want them to be like productive. We want them to be used. We want them to be moving around the system. Uh and if that's blocked by these kind of user user experience difficulties, um that's not

going to happen. So Yield Seeker tries to solve this. Our agents uh do stuff for you. So the first agent we created earns you yield on your Sable coins. Um and it's similar to having money in a savings account except everything's automatic. Um the beauty comes in that you talk to your agents in English. You give it an investment thesis. You can give it a personality. Um and then they go off do all the work, analyze the data and execute autonomously and safely. And then they come back to you and explain in English. So you can iterate on your on, you know, how you want to be investing, what kind of actions you want it to be taking all again in English. So a quick demo of our, you know, what our UI looks like. You can see my agent's been moving around constantly. It never

sleeps. You know, it does multiple things a day. Um, this is huge. And then you could talk to it in English. You know, it explains it back. You could have a real conversation with it like you would with a money like a manager, someone that's managing your money. Um, and then you could ask it to like, I don't like this. Let's change it. And and does that make sense to you? Yes. Do it. And it does it for you. And you can go ahead and see it in a nice UI as well if you want to. Cool. So, we started ETH Global last year at a hackathon and now we're managing 5,000 agents and $1.5 million. And one of our kind of secret sources or the magic is agent wallet kit which we created to make sure all the agents onchain actions are verifiably something

we want them to do. So how do we make sure agents aren't doing something accidentally wrong? Um and with that we have the power to identify all of DeFi and that's why we're raising our first round of funding um to really you know give everyone superpowers. Uh thank you. Come talk to me or my co-founder Arthur Hi, can you give us a big background a bit of background on your on yourselves? Um, sure. Yeah, that would be good to know. >> So, I mean, we went to uni together. We both started computing back in 2013. Uh, so I've been working in AI startups since then and I joined in crypto in 2020. Um, Arthur's been working in a bunch of other startups. He joined crypto in about 2016. Uh, been ETH native since then. Um, and then we've

been working together in crypto since 2020, mostly on NFT projects. This is actually our first non-NFT project. Um, and we did it because from my experience in I was working in AI before LLMs like you know back in the old days with with deep mod deep learning models. Uh, and so when LLMs came like we saw the power of like using deep learning models with LLMs uh to make DeFi easy for someone like me who kind of was always scared of DeFi. Um, and that's how we started and now you know we give that power to everyone. And you know, even just at this pitch competition, we've seen a bunch of like several eel type of products. Yes. Uh eel bearing products.

Um I'm curious where do you think your differentiator is? >> Yeah, totally. Oh, whoops. Uh it's it's kind of seems straightforward to me now why there's there are so many yield products. It's like the like I said, it's like the easiest part of this problem, but the problem is massive, right? Like all of DeFi has these problems. The yield version is just like the entry point for companies like us and and and Zipi and others um to get there. And you know the the real interesting thing is like the journey into DeFi like we're all going to take different paths. So I don't doesn't really feel like differentiation is necessary now like our for example the thing that we do differently from everyone else is our agent wallet kit right? So we make sure that the agents actions on chain can never be um it can

never do anything accidentally. It can never uh put the wrong parameter in a function or call you know send the money to the wrong place like that's verifiably onchain. Um but others you know they they do a lot more like offchain like um monitoring of the vault and stuff like that and others are more like marketing like it's all uh everyone's taken different approaches and I think in yield it doesn't make that much of a difference but as we move out into more um more activities in DeFi like it will you know I think in a year all of our companies will look very different to each other they just look similar right now.

Um so basically after users spend money on gas um on sleep edge on paying you a fee how much does he actually net versus just money in DC world fun to answer because we charge no fees users don't pay any gas they literally just they have to pay gas to transfer money to the agent but after that point there is no fee right like the the vaults will have a fee but we never show them when we show them an AP API. It's we recalculate that API, right? Like it's calculated on your balance every day. Um, and we don't charge a fee yet. We might one day, but we're more excited about building other types of agents and doing, you know, activities that were worthy of a fee.

Um, so they get the role like they get the best. >> Where's the business then? >> Well, so the other agent, so we're planning to create other types of agents, right? So agents that do maybe rebalancing portfolios or trading pers or you know prediction block like wherever a user can talk to an agent in English describe a thesis like you know I want to switch between gold and and um oil based on this uh do it for me you know that's a bit you obviously they need to talk say more than that but something like that and then it feels like a lot more viable to make money in those markets and then the yield is like the you know the savings account at a bank um where it's like the case product you come and put money here but use our exciting products and hopefully we can

make more money there but charging a fee is always an option of course but we just don't do it at the moment >> so how much of a AI makes decisions versus like a rulebased >> yeah interesting so that we've tried multiple things at the moment the way it work so in the start it was like all AI um we found that to be really inefficient mo expensive but also um like too unpredictable and then explaining it back to the users became very hard. So because one AI would then just try to explain what another AI did and like often the reasoning wasn't very sound. Um so what we do now is like if you think about the the inputs and the outputs those that's where the LLMs come in like they turn the parameters for the

model into stuff the user can understand and then the model runs and it's very easy to explain. Um but being able to like input in English and then agent output in English and then iterate like that means that the complex all the complexity underneath um the agent can explain to the user and it's it's like an interface layer really. >> Thank you. >> Thank you. >> Amazing. Thank you so much. >> Thanks. Thank you. >> We're going to get ready for our next pitch and for this one like to bring on Radius and to tell you all about Radius. Please welcome AJ on stage. >> Hello. Hello, I'm AJ from Radius. We are building onchain repo platform um for institutions to secure funding. Repo is

existing institutional lending protocol simply borrowing cash by selling safe and liquid collateral like a US treasuries with a promise to buy them back as repurchase agreements. Um in institutions like hedge funds have depended on this funding protocol to scale for decades. But there is um a massive mismatch there. Small funds with less than1 billion dollars aum make up 85% of the market are controlling hundreds of billions of dollars in assets. But they are completely being um shut out of this funding market because dealer banks have balance sheet constraints from

regulation. So big clients are preferred leaving small funds uh with crazy pricing. Also new SEC clearing mandate for repo trade coming up next year requires a prohibitive $100 million deposit. So they urgently need new funding routes. So that's why we bring repo protocol on chain. We fix these problems with direct P2P trading with non-bank institutions. So um with the bankless P2P funding routes by um bypassing dealer banks and settling instantly via our smart contracts we eliminate this balance

balance sheet um costs and permanently uh fix price discrimination. Also we bring ownership transfer on chain meaning cash lenders uh own the collateral collateral from day one. So uh keeping them completely safe from bankruptcy estate on hedge fund side. So um on our platform hedge funds can see a list of cash lenders with maturity haircut and interest rates and collateral that they accept.

In fact smart contracts handle these figures and margin call. All participants can trust the system completely. So um hedge funds provide the existing tokenized money market funds like bele from black rockck as collateral to lower their net borrowing costs. Cash lenders lend the stable coins based on the conditions matched by um our smart contracts. These transactions are settled via DVP process onchain meaning that automistity for cash and collateral exchange is guaranteed.

So um we are doing beta service for um bunch of institutions and hedge funds right now. So come and talk to me um after this pitch. Thank thanks >> um actually can you walk me through the flow of um you know the escrow transfer? So it's like I have a tokenized T bill. I lock it in the contract. The counter and the count maybe it's like before the counterparty and I agree on a rate and duration, we lock in the contract and then the stable coins are deposited within some number of blocks and then

settled or like how because it seems like there's some asynchrony here but maybe I'm missing something. >> I see there is a smart contract matching engine. Yeah. that um cash lenders and hedge funds allows these smart contracts to move their assets stable coins and collateral at the same time. So if these two are kind of agreed that on this the exact terms this matching engine um handle this u the settlement process at the same time. >> Okay. So >> sorry. So, how do as a user do I is it do I should I think of it as a full order book that's transparent or should I think of it more as like I post a private bid and then it only it's only

matched if like some condition holds like >> yeah you can think of as a order book uh order book system as well so everyone can see which assets or uh sale coins collaterals are actually on the market >> and do you have um different order types though like if I'm an iceberg like I want to send I want to do 100 million trade versus like a million or less like how are how do you think about that? >> We we actually thought about that but we haven't um um implemented that features but in the future maybe. >> Cool. Thanks.

>> Thank you. >> Sorry. >> Hi. What kind of sizes of target funds are you looking at right now? And kind of just talk about the traction you have or it sounds like you haven't launched yet. Is that right? Or you're launching there now? >> We just started a beta service. So the number is not that big but um the the size of fund is doesn't matter because um as long as these two counterparties are agreed on fixed terms the trade can be done. >> So what's the size average uh you expecting from >> you mean the hedge fund?

>> Yeah just for around 100 million or >> no no not the hedge fund itself but the transaction itself that's taking place in radius >> transaction. Uh okay. Um the tradition traditional finance there will be like a like a hundreds of trillion but I I uh in real practice hundreds of million I suppose transition >> like how how much is a hedge fund borrowing on on your platform on average? >> It it depends on the cash lender side like institutions treasury teams.

>> Yeah. But when you're just starting out like what do you think the size will be like? um tens or hundreds million I suppose. >> Okay, you're gonna start out with that. Okay, >> I see. >> So, and where is your spread versus uh like a third party repo and more for fixed rate? >> Sorry, what was that? >> So, where's your spread between like a third party repo and like a more for fixed rate? >> Okay, so there is no um trading fee um that these parties paid to us. Our our business model would be like a post trade uh system like a collateral management or reporting generation. We can um charge those kind of services but no trading fee on our trade.

>> Yeah. And in terms of like go to market uh essentially you are a marketplace. So how do you establish those relationship? What geographies? Uh how do you build those >> kind of both sides demand and and supply? >> Yeah, actually we're targeting the Asian market. uh since like there's a lot of tradition from the cross border uh repo trade actually so we're targeting institutions in Asian and then also doing some um business in US thank you thanks >> amazing thank you >> next up we're going to bring on unlink and to tell you all about unlink please welcome Paul Henry Hello, my name is Paul Henry Cashfaz.

I'm the founder and CEO of Unlink. Last year, there's been more than 100 kidnappings and attacks on people because of the publicity of onchain wealth. And in the age of AI, this is only going to get worse. With crypto, we've built the best financial rails ever invented. They're faster, they're cheaper, they're global, they run 247, they're fully programmable, which means that companies could save millions of dollars a year on fees. So why aren't companies actually using this at scale? Why isn't everybody using blockchains?

The reason is, and one of the reason is because every single thing you do on chain leaks. your payroll, your treasuries, your B2B payments. Every single trade you make leaks on chain forever. Um, which basically is a non-starter for companies and for most of them, it's not even legal to use blockchains. I've spent the last two years working at one of the leading privacy preserving protocol. I was a senior protocol engineer at Polygon building Maiden. And at Maiden I talked to hundreds of companies wanting to bring their funds on chain and they needed privacy and compliance and basically they told me every single time the same things they needed privacy on existing chains. They

wanted to use Ethereum base Salana etc. They didn't want a new chain and then the second thing is that they needed full compliance that their legal teams would sign off on. And so I decided to leave and this is exactly what I built. We're building Unlink, which is the privacy and compliance infrastructure for companies moving money on chain. In only three lines of code, you can uh use any blockchain fully privately. You can own, receive, send, use DeFi privately on any existing chain on Ethereum, on base, on Polygon, on Monad on ARC um and many, many more. And it's not a concept.

We've had hundreds of applications built, thousands of developers. And during the hackathon, we've had hundreds of people buy our merch fully onchain, fully privately. And this is not a concept anymore. Um, we are already shipping with some of the best teams in the industry. We're already had closed deals with Monad, with Oiler, with Open Fort and many, many more. Actually, we're closing deals at a very rapid rate. We're growing uh and we're closing new deals every single week. We've built a team of experts uh that have built this in the past at top companies like Polygon, Tartnet, the Eterm Foundation, E Global, Microsoft, and more. Uh and we're super excited.

The team is growing very fast. So with Unlink, we're very excited to bring every single company on chain. And if this resonates with you, come talk to us privately. Thank you. Hi, thank you for this. Um, can you just talk about more how the architecture works and how you're working with projects like Monad? >> Oh yeah. So basically Unlink is a smart contract that you can deploy on any chain >> EVM or >> it's a EVM smart contract but um it's fully chain agnostic so you could deploy it on pretty much any chain. So right now we've deployed on all these EVMs but we're looking into supporting uh Salana

Stellar etc based on demand. >> How do how how do you work with monad like how do they integrate or work with you? >> Yeah basically we deploy the smart contract on monad and then any application can build on monad fully privately. So you do um basically you do here um con on link create client and then in the brackets instead of ethereum you say monad and then this is a fully private account on monad and then basically with monad what they do is that they put us in their docs and then now it means that anybody can use monad fully privately by default.

>> Yeah. So what is under the hood? Is it like a ZK, FHE, TE, secretive addresses? And what is it like in terms of compliant? What does it means? It's not being able to view keys or anything else like a screening. >> Yeah. >> So every single transaction is screened. So we're working with a company called Predicate that wraps basically TRM, elliptic chain analysis, and more. Um so we make sure that no sanctioned entity we do AML um anti-terrorism financing depending on the client sometimes we can do KYC KYB um their internal blacklist so it's fully programmable compliance um and then how it works we've looked at

all the different ways of doing privacy tees IO ZK FHE etc and the most efficient and the most trusted and tried is zero knowledge and that's what we've built we've built Starknet, we've built Polygon, we've built Maiden, um, PhDs in cryptography in the team. So, we're really experts in ZK. Um, and we believe it's the most efficient, the most fast, uh, to be able to do privacy on chain. >> Thank you. >> Hey, thanks. Um so I one question I have is you know when you think about these kind of like payment kit type of embedded uh you know JavaScript libraries or libraries that you want to

get distributed how are you thinking about distribution? How are you thinking about distribution of agents versus human users? And how do you think about like hey you know you had a bunch of fintexs listed here right? How do you get them to actually integrate um this like how do you think that? So for example, we we run a hackathon uh not too long ago and one um of the judges of the hackathon that's really interested and that's becoming an angel is the co-founder of Venmo for example. Um or for example open for they're doing a lot of stuff with agents and so our stuff is fully compatible with any agent. Um so any agent could like just write these three lines of code and send a transaction on chain. Um and in companies um we really think that this is obviously the future.

So we make it fully agent ready. So we have support for X42, MP and all these different platform and standards like uh we're doing something at the hackathon e global um with arc with nano pay uh and dynamic. So we're doing this joint track uh and so I think that's going to be super exciting. So yeah, it's fully compatible. >> Yeah. Yeah. I just meant more like how do you think about you know Stripe was successful because they added three lines of JavaScript but then they really went and like got a ton of people who had small companies to use it. So how do you think about your distribution model for for that? Yeah.

>> Yeah. I think um most companies if they want to integrate blockchains they will need compliance. They will need privacy. So this has been a very very big hurdle and there were ways to do privacy but it was extremely slow extremely expensive like non-compliance. So like these companies couldn't really integrate it and it was super hard to understand like I don't know Zcash or tornado or all these things. Uh and so we make it like super fast, super cheap, super easy to integrate and then basically we do a bunch of distribution with our partners. So Monut pushes us to all their institutions and to all their companies open for to all their developers. Um Oiler is basically bringing their curators, their institutions, their whales to be able and we integrate in

their front end to be able to bring privacy to their users. So these are all like clients and distribution partners. And we're closing a lot more. Like we have I need to go back to work because the pipeline is full. >> Amazing. That's all the time we have. Thank you. Thanks so much. >> All right, three more pitches to go for our next team. Let's bring on Dolphin to tell you all about Dolphins. Please welcome Titi. >> Hello. Hi everyone. I'm Titi and I'm the founder of Dolphins. We are building payment infrastructure for the future of work.

This is a typical invoice that we see in our space. Three wallet addresses at the bottom of an invoice with some instructions to not pay to an invoice that was sent earlier with the wallet address from earlier. So the client that gets this says, "Okay, let me process this." So she sends them some crypto. Then a second invoice comes in and now the payment stack expands and then a third and then a fourth. As the team gets added, the client ends up doing AC, Bitcoin transfer, ETH transfer, so many of these. And guess what? There are then compliance questions that she also needs to answer. So we thought um what if the client never has to do any of these?

What if the team never have to think about what I'm going to ask a client? So we build dolphins with just one wire in and finite choices out. The client transfers a single transfer. Dolphins routes, converts, settles everything compliantly. So a contractor in India receives INR and ETH. A contractor in Brazil receives uh Brazilian real USDC and ETH. Every contractor according to what they want without doing any additional work. By doing so, we actually enabled non-crypto users to also come to the platform. We rolled this out and then surprise surprise, we realized that we actually

can use our rails for a different kind of monetization by uh offering API rails and web hooks. So, blockchain rails uh can settle very fast and the adoption has actually forced regulators to provide some clarity. But the number I'm very very interested in is 1.57 billion people around the world working independently as freelancers today. And this number is going to be 2.3 billion by 2030. That's $1.3 trillion paid to these people. But every payment platform is built employees first, fiat first and bank oriented. We are building web 300 under the hood, web 2 on the surface, contractor first, cryptonative.

about 1.3 trillion is paid to uh this workforce and even if we conservatively look at.1% of this we tapping into $39 million revenue potential we make money when money moves 3% transaction fees and u infrastructure fees for licensing and APIs we launched uh we're live and we processing across eight countries we have two API clients that have already that are already active and two more uh integrating with us I have spent more than 20 years with fintech and 13 years dealing with this problem. Chris, my co-founder, has built products from zero to uh to production for two startups.

And our team is actually paid through Dolphins. So come and use Dolphins to pay your team to stay cryptonative and to build your business without borders, banks or BS. Thank you. >> Thanks. Um, so I I'm curious how you think about the sort of competitive landscape here, you know, like the open FXs of the world versus this and and sort of how do you think the fee fees will stay kind of stable versus having margin compression over time? like what what do you view because I you know there's a lot of people kind of everyone knows this is a market so I'm just kind of curious how you see that evolving how do you see differentiators things

>> right the two kind of cost that comes into actually crossber payments when it comes to employer employer and contractor pay one of the cost is born by the employer averages around 6% to send uh money across border the other cost is a cost that a contractor actually ends up uh taking because of currency conversion and all of that. If you net that, it's about 13% a typical contractor uh loses in this. I quantified that and that's about $90 billion just lost into hidden fees and all of these different things. Because we can stay cryptonative, we actually can use a lot of uh different uh ideas.

My secret source is actually the workflows that we have built contracts, escrow, all of these. We bring them cryptonative and we give the client a a way to say can I generate yield on this while it's sitting in that escro account. So doing all of that we are able to offset a lot of fees that we take from the client and also give non-crypto clients a very easy way to get some benefit from uh the crypto industry. In fact my first three clients were completely non-crypto. They didn't want to touch crypto at all. Today they process a lot of their um of their payroll in crypto which is really fantastic.

>> And um you know how do you view auxiliary products beyond the initial sort of um kind of setup you know things like you know a lot of different jurisdictions have like earned wage access type of lending stuff like that like how do you view that in your road map? >> The limit there's no limit it's limitless. So there's earned wage access which is a very big pain point for contractors. You've earned for one month you invoice a client you have to wait another month to actually get that payment affects your cash flow a lot. So earned wage access is definitely something that we are exploring. Um and we already have spoken with some credit unions to actually do that for collection. That is one. The other one

that I'm more excited about that we're going to roll out by the end of this year is uh cash out. So we are integrating with Moneygram so that uh contractors can come to our platform, split their payments, allocate a portion of their uh income into a wallet and they can cash out through that wallet anywhere around the world. And once we do that, we would be able to use the same concept to also do cash in. Especially for gig workers who get paid daily, cash in is still a very uh lucrative uh rail. >> Hi. I think maybe just picking back off of what Terrone was asking about. I'm curious about the profile of your customers as in uh the B2B customers primarily on you know how many contracts

do they have? How many geog uh jurisdictions are you serving? Uh do you need licenses for those serving those jurisdictions right and and what is like the sweet spot for Dolphin versus you know others we're seeing in the space where it's like you know could be as big as deal targeting like international payments and payroll. Um yeah >> uh we have been experimenting with our markets and last year we uh early this year we realized that our markets is really in Latin America, Africa and South Asia particularly India and I think India also because I do have a deep network there. So those are the ones we're focusing on to restrict spending our of our marketing dollars.

But the interesting thing is that non-crypto users are more comfortable using our platform because of the contract feature that we have where they don't have to go and create another contract. They can just come and use it. It's completely web two. You look it looks like a form underneath it's all the niceness of uh web 3 which they don't need to know. So that is one. And then what we've also started experimenting is we take uh our crypto rails and we are going to smaller payroll companies and we're telling them hey you guys are never going to build this so take our rails and then use it and that actually is giving us a a a nice pipeline of people who wanting to pilot it and actually that was the reason why we built our API and web hooks to give payroll companies and in

into the crypto rails. >> That's all the time we have. Thank you so much. Two more pitches to go. Let's welcome our second last team, Frames. And to tell you all about Frames, please welcome Luish. >> Hey, hey, hey. Good morning, everyone. I'm Luiz. I'm one of the co-founders of Frames and we help your AI agents access the best data on the internet. We've been around for about 12 months. We started at the NET global hackathon back in May 20 2025 and since then we learned a lot. We've been deeply embedded into into the agentic payment space and this is all I'm going to talk about today.

AI agents in harnesses and models they're they're getting great. Yesterday we we we saw Fable 5 launch. Uh it's a great model by the way but the problem is when they try to reach for the paid data providers which are generally the best ones they get stuck. And the reason why they get stuck is that they don't have an easy way to pay for those things. So that leaves agent operators, the humans behind the agents, uh to integrate each provider individually. This means API keys, accounts, different payment rails, budgets. It's hard. So we solved that with frames, a one model to access the entire payw wall internet. It's called a composite model because it

combines several components. frontier models plus a vast tool catalog that's growing of more than 4,000 tools and we own the loop to try each individual tool, curate them, rank them to make sure that in the end you get your outcome in just one bill. It works across all those those clients that you yall use OpenAI codeex cloud code etc. But the benefit of this is also we support OpenAI compatible endpoints meaning that you can hit the API for inference.

So let me give you an example. You hit frames with one request. We find that there's a data source that's paywalled the cheapest reliable source. We pay for it on your behalf and then we verify if the outcome matches what you want. If not you you don't get charged. If yes, yeah. So we've seen several use cases from GTM research, financial analysis, data enrichment being used using our model. We launched initially in February this year and since then we've seen more than 400,000 requests, meaning microp payments across our tool catalog.

Our insight comes from all of this data. Agents need better data. We're frames a one model to access the whole paywalled internet. You can find me around. Thank you. >> I was wondering uh who pays first, the agent developer, the merchant or the user that delegating the purchases? >> Who pays first in your first?

>> Yeah. Yeah. the agent, developer, merchant or the user. >> Yeah, we do we do support a subscription model at the moment. So the user subscribes to our to our product and then the independent tools we pay them ourselves and you only get charged if the tool actually matches your your request. >> Um and so actually I'm sort of curious how you think about sort of uh tool routing and token routing and and monetization around that. So, you know, right now you're driving in theory you're driving order flow to these subscription models, right? Potentially you'll get referrals or or or some type of monetization that way, but there's also other monetization it looked like

in your architecture diagram of, you know, you're doing some token routing across the different models. You're also doing some type of like tool routing of like, oh, this tool took too long. I'm going to cut it and try another route. So, how much do you think you'll be able to monetize on the optimization side of like the token writing side versus this kind of like referral kind of affiliate marketing side? >> Um, you mean like different models, how do we pick them and uh the different tools and >> Yeah. Yeah. And then like how do you view the relative monetization of the two like which one in the long run will be how you make money? >> So, right now it's pretty simple. We charge uh a margin on top of each tool call we make. Uh in terms of model we keep it simple for now like we keep uh

we use for example for the base tier or like the basic the cheapest one we use a very cheap model like I think it's Google Gemini or Nano I cannot recall the model right now but but then for the the second best one we use uh Opus 4.8 98 I think and yeah so on and so forth like we're not really routing the models yet okay >> right now it's mostly like finding the tools based on the data we've uh we've aggregated over the past couple >> the reason I'm asking is more like once you have enough workflows and then you start >> being able to pay less to route roughly the same access then you can start charging users on how much you save them right like I saved you 80% by not just using cloud 10 times in a row >> right so so do you view that as like a

longer term kind way to scale. >> Yeah, potentially for sure. Um >> but I I cannot really expend more on that, but it's a great idea. Yeah, sure. Yeah. >> Maybe just quick question on I mean it sounds like you guys already have found some use cases and research and other areas. Do you want to go vertical and deep in certain areas and segments? How do you think about capturing the next leg of like user base and sticky user base? >> That's a great question. I think um well potentially for sure like trying to to narrow a bit down uh that's and you know like we've noticed from the data that people people are trying to get data from you know Twitter or you know doing deep research stuff like that so

it makes sense for us to narrow down but we also noticed that the most of the users just plug in the our model as a second brain to their existing harness so it reaches out for like those longtail use cases that we don't really know yet. Uh so Not sure if that really answers your question, but we're keeping it open-ended for now. Uh, but potentially, yeah, we can we can narrow down if if we have a group of users. >> Yeah, maybe just just one more. So, every pay wall is basically a different system, different financial model that you need to kind of integrate to. So, how do you manage refunds in that?

>> Refunds. >> Yeah, >> we we don't do refunds because I mean, what do you mean by refunds? like if you want to get refunded by the subscription or >> Yeah. Yeah. Yeah. stuff like that because it's every system is different and then refund is like a one integrated system within your product. So how do you call each different system? >> So so ideally we absorb a lot of the risk from trying different tools. Um and you only get charged if the that actually matches the outcome. If you're not happy with it, you can ask for a refund and you get your your money back.

>> For what it's worth, that's why I asked about monetizing by minimizing tool costs. >> Okay, gotcha. >> All right, that's all the time we have. Thank you so much. >> All right, and we are ready for our last pitch for today. So, please welcome M to tell you all about pulse pay play. Tomorrow afternoon at 3 p.m., the World Cup will kick off in front of the largest live audience on Earth. For an inplay event like a penalty kick, there's a 10-second window where the entire stadium is holding their breath.

And right now, there's no way to place an onchain prediction on that moment. So, that's what we're building, the settlement layer for markets that move in seconds. So, why doesn't this already exist? Well, you might assume it has something to do with speed. Maybe the chains just are not fast enough. But the real issue is how the infrastructure works under the hood. They match your trade offchain and then settle it onchain a few moments later. Now, during that gap, a trade you've already won can just disappear through a mechanism called a ghost fill. That gap is exactly why no one has reliably built a second by- second market like for that penalty kick. So, that's what we're fixing with Pulse Play. You open a session, you place your predictions, and you cash out without the funds ever leaving your wallet. There's no gas on any single

prediction. Just one transaction to open the session and one transaction to close it. Every prediction is final at the moment you make it because it's a signed agreement with the pool, not a promise to settle up later. So there's no gap, no ghost, purely by design. So how do we actually do this? It all comes down to a technology called state channels. Our industry wrote these off about 10 years ago because they kept trying to use them as a general execution layer, but they're not good at that. Now, as an accounting layer for markets that move in seconds, they're actually the perfect fit. And the lightning network has been proving this with Bitcoin for years. So, sports is where we're going to start, but the layer underneath is what actually unlocks these short duration markets. The timing here is not an accident at all. Polar market has been

trading at a high volume for a really long time and Kali traded at $40 billion last year. Nearly 90% of that was on sports, especially since regulators have opened up the door. But the one thing that these big players still can't give you is that live fast inplay moment. So we're not really competing with Polar Market and Kaji. We're building the layer that they don't have yet and it's not even on the horizon. Now a quick word on the team. I'm one of the technical co-founders and I just wrapped up shipping a regulated RWA token in Europe. So I'm full-time on this now alongside my co-founder Liam who's on architecture and Tyler who's on growth.

We won the very first yellow network accelerator program and we were winners at the ETH global hack money hackathon where this project actually started. Now look, we're not up here claiming that we're smarter than the team at Polar Market. Our bet is simpler than that. Being this deep on the right primitive this early will compound faster than anyone else can catch up to and our infrastructure is powering new types of markets that are complimentary to what they already offer. So here's where we are. The hard part is already built and we go live at the World Cup this week. We're super early. We're self-funded and we're open to the right strategic partners and angel investors. So, if that sounds like you or you just want to see how this works, then come find us because we're happy to chat.

>> Hi. Uh this is super cool. I'm curious to learn about how the architecture works um under the hood. >> Awesome. So, we're powered mainly by something called a state channel and we're closely aligned with a team called the yellow network. This is something that is a bit different from a rollup or an optimistic rollup, anything like that. Um, it's actually kind of like an off-chain model where you have a stack of signatures that build up over time to keep track of your transactions. So, how it works is you deposit your funds into a custody contract on any EVM chain. Those funds get sent over to the other network and then you can transact with web two speeds sub 10 millisecond latency for as often as you want. And then once you have that stack of signatures, eventually everything comes back on chain when you want to withdraw.

And then you guys will be a standalone platform or you're looking to embed yourself say for the World Cup in a live streaming or any kind of platform like that. >> Great question. So we're seeing the World Cup for the next six weeks as a big data collector. We're going to use that to see how the platform scales at, you know, a large number of concurrent users and then once that's done, we have a number of different verticals we can go into. So in the sports vertical this is actually really good for iterative games like baseball and cricket. And then aside from that we have these rolling markets like what polymer market does with Bitcoin for every 5 minutes.

We can shrink that down to actually one minute. Um so there's that and then there's also live streamer engagement. Uh streamers are always interacting with uh their their fans and people pay them to get shout outs. So this is another way for them to interact by proposing different things that can be done etc etc. So there's a number of different verticals this can expand into, but we're really focused on making sure the infrastructure works. It works at scale and can actually power this new type of prediction market that is very very shortlived. >> So yeah, I believe that like DraftKings are doing it offchain pretty successfully. What would drive people to bet on things like that on chain?

>> Great question. So we are focused on the web 3 market. Uh the web 2 market with DraftKings, Bet 365, FanDuel. Uh they're sports books. So they do offer these types of markets for their sports books. You have to do comprehensive KYC with them and uh you can use fiat but you can't exactly use your own USDC or your own cryptocurrency in a way that's self-custodial. So a big deal for us is the fact that your funds never actually leave your wallet until the moment you place your bet. It's akin to let's say you're at a casino and uh you exchange your money for poker chips. You can exchange those poker chips back for cash at any point in time. It's the equivalent of money inside there. when you're at the table and you're sitting down, it's still in your hand, but the moment you place a bet and you put it

into the pot, that's the only time you actually relinquish control. And that's pretty much what we're doing. We're using the same exact model. >> Thanks. >> Um, so actually I'm curious how you deal with um oracle settlement uh within kind of the channels because like I'm assuming that you know we have sort of preunded state channel. Um, I have an oracle that's hopefully f sufficiently fast, but like I could imagine there's some timing game of like the last Oracle update didn't come. I hit withdraw and settlement and actually the market resolves the other way. So, how do you how do you how do you think about interlacing trades versus Oracle updates and like what are kind of like the safeguards there?

>> So, that's a good question because the Oracle isn't actually onchain. We're closely aligned with a data feed provider called TXOddS and they're giving us extremely low latency feeds. Now, I'll be honest. If you're in the stadium venue, that's the best place to be because your latency is your eyes, right? So, it's 200 milliseconds. If you're away from the venue, then we're considering your phone your second screen. So, let's say, for example, you're looking at a soccer match. Um, you're going to see the update on your phone before you see it on the screen itself. So, that's the that's what we're dealing with. Like, we're making sure that you get the fastest updates, you can access these things, and the event feed that we're getting from TX Odds, it's actually being curated for us specifically for these markets. So they didn't offer these data this data

beforehand, but they're getting it specifically for the World Cup for us. So we can collect this data and then prove that it works at scale. Do you do you think that there will be other Oracle providers for low latency feeds like this? Because like obviously the World Cup's like one big event for this and admittedly one of the biggest betting events at overall. But like you know if you want to scale this, you're going to actually need to have the same fidelity everywhere. >> Yes. So I'll admit that's one of the n that that's one of the strategic partners we're looking for. But we have been in talks with Golder. We're closely aligned with TXOdd and then Sports Radar and a number of other parties are on our radar as well to to get closely aligned with. So I mean like I said we're really

using the next six weeks as data collection. We're proving product market fit because we have some test users but we want to make sure that we refine that perfectly and then once that's done then we can put the 25k 50k bill per month to pay for like that data feed. Thanks. >> Amazing. Thank you so much. Those were our 14 incredible startups that told you about what they're up to. Next up, our judges are going to deliberate on who the winner is going to be for the best pitch. So, we're going to come back in a few minutes and we'll tell you what the best pitch is and which the best team won.

In less than 10 minutes, we'll be back here. [applause]