ETHIstanbul Conference 2025 | Permissioned Shards Meet Public DeFi | Nicky Chalabi, Pelagos

Transcript

[Music] Okay. Okay. Thank you so much for joining the talk. My name is Nikki and I am the CBO and co-founder of Pelos Network. I'm sure guys you haven't heard about us because we're quite a new project. But I will tell you today what do we do at Pelos? What do we want to achieve and where we're heading to actually what kind of problems does Paloo solve? Right? Like it's not just another chain. It's not just another project. Our aim actually to kind of bring this atomic execution to the space and to ensure that there is kind of communication uh between the chains and also we'll be exploring like what can pelos bring to the institutional

landscape especially now we've seen like bunch of for example projects such as circle such as stripe launching their own like institutional grade chains and essentially how pelgos can fit into it. So we're by no means like specifically servicing those sorts of solutions but we are also like providing like we're also able to provide the solutions for them as well. Oh, give me a sec. Okay. So, generally speaking, right, if you've been a lot in the space, like one key thing you notice, especially like if you played a lot with D5, for example, you see that uh you have bunch of chains, you have like bunch of L1s, L2s, and I personally do believe that there will be more and more chains coming into the

play. And um I mean my thesis is personal thesis is that like we will not get any more maybe general purpose chains. most probably we will be getting like more purpose specific chains so it will be purpose specific L1's purpose specific L2s and you see that for example if I want to for example do something like if I have my funds let's say on optimism and I want to do something on arbitrum I have to like utilize bridges and despite that like the liquidity is still fragmented right like you can like basically if I have my liquidity on arbitrum it's stuck on orbit room if I have my liquidity on optimism it's stuck on optimism nothing changes unless I take an action and this is also problematic for uh DAPs as well

because in the end of the day like DAPs when they deploy they go chain by chain they have to deploy chain by chain and it kind of erodess this multi-chain experience because like your for example liquidity on chain A is stuck on chain A right like and you have to kind of incentivize to attract that liquidity to that chain and the solutions that we currently have like you have bridges and messaging systems that are asynchronous and create lag risks. So I mean of course there's still like there are good solid bridges solutions that help you move your actually assets but there's still like they still do introduce certain latency and certain like risks and the same applies actually for the

intent like networks right so like their intent frameworks I'm big fan of them because they have managed to resolve certain UX issues that we're currently facing but still they do introduce those latency like issues that might a problem for like certain players right if we're talking about high frequency trading and so on and then if we look at the institutions especially the chains that are currently launching like that are geared towards institutional adoption like one key thing you have to take into account I mean especially if you go to crypto Twitter you see like for example a lot of people commenting like oh they should have been an L2 or oh like they should have used this chain or that chain you're like come on like have you

worked in a real work like job Like because when you work with institutions you see that they it's not them deciding like the requirements. It's actually regulator telling them like we want this right like they say like okay if you want to offer this product you have to actually go and ask the regulator first. So you're not like just doing it out of like your desire. And um what Pelos is is actually an atomic crosschain execution layer. So what does it mean? So it means that like I have a transaction let's say I want to like I have a pangu on Solana and I want to send like a convert it to USDC on Ethereum like I send this transaction once and what Pelos does it just like

simulates the entire route and like basically like if all the constraints are cleared like the simulation happens actually in parallel then the transaction passes so you don't have any bridges intent network so we remove that kind of complexity from from the UX and essentially as mentioned like the new rails are coming like such as circle arc L1 you have stripes uh tempo L1 and we are able to kind of like you know take those price L1s that are coming into the space and kind of be the atomic gateway for them to access the public defy.

So what does atomic actually means here? So that's basically one of the questions we need to ask uh ask ourselves. So uh generally speaking as mentioned right let's say you have a transaction you submit and it can be as multi-chain as possible. you can say like I want to swap here um I don't know provide liquidity there like uh borrow somewhere else and uh what happens then is that like this whole transaction is taken as one and then like you have this entire route simulated like which happens in parallel and if all the constraints are meant like uh for example when it comes to gas to slipage uh we have this all or nothing semantics so the transactions to

each chain RTSS signed only if the entire bundle actually meets the constraints otherwise nothing leaves Pelos and then you have this like one signature one decision slot in less than a second. So then each leg uh settles on its chain per normal one key thing to take into account. So internal decision is subsecond final settlement then follows each destination's chain block time and finality. So when it comes to the final sentiment on a chain like we cannot influence that but what we can influence is actually what's happening on pelos.

Um so let's come back to intent networks because we do get confused like sometimes with intent networks like people ask us okay like are you guys using intents or what kind of bridge are you using in the back and so on. Um generally speaking like um there are a bunch of solutions right like when it comes to intent uh like networks and so on like you have the likes of cow protocol for example you have unis swap x and so on like which I'm like a big user as well but um the thing is that about like they don't still provide you with uh crosschain uh execution layer atomicity so you still depend on the solvers and the solvers still need to find for example the solution quads for

you and sometimes like that can take time and during that time your position might move for example. The other good example of intense centric solution is anoma which uh verifies coordinates intense but final in uh like execution it still lands chain by by chain. there is no like a uh crosschain execution layer atomicity and um in case of pelagos we don't just like route intents or transactions we actually decide and cosign the entire multi-chain transaction as one unit so how pelos actually works so here we introduced the concept of app chains when we're talking about app chains you

have to speak uh think a little bit about it differently so app chains are not the ones that you are used to maybe in a classic sense of it. It's not like for example dimension rollups or something some other solutions that exist there where you have like app chain as a service and so on. In Pelos app chains you have to think of it as um like a micros service. So what does it mean? So it's a dockerized program shipped as an image and then what happens is that validators they pull it they run it they upgrade it for you like they do everything in the back and um that's basically what you have to think of and what does it mean to be an execution layer right so it means that with your app chain you can be your own

execution layer you can decide your rules you can decide your like for example gas fees you can decide everything like your token whatever your economy as you want. So you are actually running your own state machine from that perspective. And then like when you submit transaction as mentioned before you have the simulation happening. So like the whole kind of thread is simulated the whole transaction is simulated in parallel. And then you can also introduce for example private ordering. So there's no meth leakage to the public uh meles. And then you have this like you know atomicity uh when it comes to decision and signing like native uh transactions are broadcast in parallel to each chain and then like

inposes okay if all the constraint constraints meet there's like a basically commit or revert. So from the perspective of the diagram it would look something like this. So you have the user or the app chain and then like there's a simulation then like there's a private ordering you have the threshold signature you go it goes to different chains and then you have the commit state. So and basically how for example could it help benefit the institutions right like now that given there are so many solutions that are are launching in the space just like targeted to launch like um institutional like institution for regulated uh chains and so on. So

basically what's really important for the enterprise level I would say clients institutions or partners like they have to be able to decide who is able to participate in their network. So what do I mean by that? They have to be able to whitelist their uh the wallets. They have to be able to actually okay KYC their customers. They have to be able to kind of you know follow the transactions and know what's going on. You cannot just like as an institution especially as a regulated entity if we're talking about a bank or a financial service provider do whatever you please because in the end of the day the regulator will come after you and take away your license and also what's really important you have to have the like audit

anchoring right like which means that the receipts that you can show okay this has happened in fact and essentially what you can do like you can be a private chart but you can still communicate with the public shard which means that you can still access what's outside of your realm like you you can decide okay which rules do I want to set for my realm and can I go outside of this realm and can I utilize the solutions outside of this realm and uh for example currently we're building like cross uh venue um perpetual decks was one of the projects in the space so essentially what you can have you can have collateral on different chains for example and you can

move that collateral very easily and uh essentially like yeah this could be very interesting especially for the professional traders for example sorry um this okay then liquidation as service this is more meant for lending and borrowing protocols so for example let's say you have a lending and borrowing protocol you say like okay um you let's say supply 100,000 worth of Ethereum and then you borrow like let's say 50,000 and worth of USDT and then at one point if Ethereum price drops dramatically low like you there's a liquidation event right like this is very important especially in the context of lending and borrowing protocols then

what you have to do you have to liquidate that ETH and get like basically the the USDT and basically you say like okay I'm I have ETH give me like all the USDT possible and you have access to all the USDT on all the pools And yeah, there's some other kind of like for example examples that you can play around. So I won't go into details with these ones, but generally speaking, it's just to show what is possible with Pelos generally speaking like when it comes to Traty and actually outside of Traty.

And there are some other examples for example unified order books uh flash flash loan bundles you can do you can do omchain yield router like payments and stable coins um gaming and creating economies you can also venture into AI and so on like you can have AI agents based on different chains communicating between each other without any need for bridges or other solutions and um essentially yes so what we're solving here and like what we're bringing uh to the space as like as a as pelos. So essentially what we're saying like okay you want atomicity you want this atomic execution we're providing that like you have this one click all or nothing execution across chains and with

decision in less than one second and uh you can precast uh you can have the broadcast simulation so you can also have the uh private ordering with that we are not a bridge and we're neither an intent router uh so like basically we actually natively move we provide the possibility for native movement and yeah so like you can also operational comfort here means that like you are your own execution layer so you can decide how do you want to set up your own logic you are basically your own state machine and you can decide what type of solution you want to run like be it private or public so yeah thank you so very much and if you would like to learn more about us we

have our website you can follow me on X and here's my uh telegram you can just message Okay. Any questions here? >> Any questions? Um, I have a quick question. Sure. You say you're not a bridge, but how do you define yourself? Do you categorize yourself as something like a clear terminology? >> Uh, yes. We just call ourselves like crosschain execution layer. We're an execution layer. >> Crosschain execution layer. Um and also so in this atom to to make the atomic swap what I don't understand is um ultimately the there have to be separate transactions on separate chains.

>> Yes. In the background this happens but like you as a user you don't see it right. You just submit it as one transaction and then like there's a simulation which basically simulates the entire route. So it simulate it checks for the gas it checks for the slippage. you know the certain set of conditions need to be met and you can define those conditions. So you can say like okay if this condition is not met you have to revert. So basically that's how it works >> right but so when these transactions are sent to the separate chains I know that happens under the hood but what happens if one of those fails >> uh if okay so the thing is that if it on it doesn't fail on pelago site >> it gets sent out like what happens on the other chains like we can't really

control outside of pelagos like because they can have different finality they can have different right uh ways uh of doing things but on pelagos if it's Everything is fine, it's out. So that's the logic that it follows. It's deterministic. Yes. >> Okay. The the automaticity >> is to send the transactions out there though. But if they once they on the the native chain that's beyond your scope, correct? You don't provide anything there. >> Correct. >> Okay. >> That's kind of hard because like they have their own each chain has its own consensus mechanism. Each chain has its own like uh like finality. Each chain has its own uh parameters, right? So we can't really control those. What we can do is like okay we we take it like as

deterministic. So yeah it worked in the simulation we checked so it goes out. >> Okay. And what's the current state? Are you in test net or >> Sure. So we've been on actual DevNet uh for a year before like um and we also like for example we had some good numbers. We actually published them recently as well like on certain like cases for reach like uh TPS of more than 100k and so on like u so you can check those out like because they're posted also on our ex um but we will be launching shortly on public test which is end of this months so and yeah we recently like actually completed our race like preede race so yeah like we're actually started also on boarding some

dabs that are kind of started building certain solutions like for example we have um like um c like developers like builders building like on us crosschain per decks and they will be using like also aail da we're working with aail there as well and then uh we have also some like other uh you know dexes that will be building also amm on top of us but yeah that's basically what's happening and we'll be releasing all that information shortly >> okay very interesting uh thank you very very much for your talk >> thank you thank Yep.

>> I'm a grant from the foundation. >> Okay. Cool. >> Okay, we'll just wait a moment for the slides to change.

Oh, it's on me. Excuse me.