Stake it ‘til You Make It: The State of Restaking — ​Bojan Antunovic | Decenter

okay hi guys so today I'm going to be holding a presentation on the state of reaking in web three so the name of the presentation was stake it till you make it and the idea behind that was that we are packaging these re these Structured produ Products these reate products as some kind of morage mortgage backed almost mortgage backed Securities where we don't even know what the underlying is so in this case there here's a tweet from from crypto Twitter that says interest you in ether 5 7 CS D USD USD USD Zer volt and mosis weake to symbiotic weate we receive token to car YT s mon maturity great yield by the way so uh we see this trend with we we come back to what weaking even is but we see this trend where we like package these Structured Products in such a way that is it is not even legible to normal normal people or even like Finance people so first of all what is weaking so uh this user asks what what can anybody explain to them what Iger does in one sentence and in layman's terms and siram canand the founder of IG layer says using IG layer or any restake product in that matter you can build arbitrary decentralized services with cryptoeconomic security we we'll go into detail what that means so even this one sentence doesn't really get to the core with what what reaking means so IG layer or weaking in general but in this particular architecture so explain it like M five so I is like a multitool for the blockchain world that helps make other blockchain services stronger and more secure what does that mean so IG allows the same security validation resources that are used like for securing the e e ethereum validation or the ethereum consensus to be used for other arbitrary offchain computations or any computations so in in this case resource multi multiple ways making the blockchain more flexible and efficient okay the ultimate goal of IG layer or restake platform is is to as I said reuse that security for validation uh and that can be used also for onchain as well as offchain operations it introduces like this concept of reaking that we mentioned which means using your already staked ether in this case uh we'll later see that any erc20 or other tokens Can could be used in such a manner so first of all we have uh the ethereum validating pool where the stakers stake their ether which is used for validating or securing the consensus of the etherum blockchain so this ether while it is utilized in this staking Paradigm and get some let's call risk-free yield or the base yield that you can get in a def environment at least in the evm ecosystem uh it can be additionally used to secure additional Services beyond that uh uh consensus uh validation or securing the consensus of ethereum so this is where I layer and other systems like symbiotic or kak come in and they say okay we can like use that stake theer and you can deposit it as collateral to be used in these AVS services which are actively validated services and they can be any arbitrary computation so it can be Bridges oracles it can be uh some defi products that depends on some offchain computation that we need some uh trust or some uh some way to verify or to slash in case of malicious behavior and so uh we have these operators The Operators are the programs or the people who run these uh sensitive offchain or onchain operations the stakers are those to Theos collateral and they are Stak in case of malicious behavior of The Operators and all this is to be used for uh any uh def apps So Def apps or onchain apps in general which combine the onchain and offchain elements okay so this is another explanation but I'll leave that for later uh so we'll go to the four levels of open Innovation as I puts it so we have uh first of all we had the Paradigm where we had one blockchain and that uh blockchain let's let's say Bitcoin in this case is the one which holds The Ledger and secures all the all the actors in the system secure the assets and the state of the blockchain uh then we we have a progression for from there where we not only securing the assets which are stored on the blockchain and the transactions and the state but also we expand it to like uh we expend the functionality to have like any program so in this case uh ether where we have like smart contracts and any like uh uh Structured Products which can be U Can can but are not confined to financial use cases uh on top of that once we saw that the true put and the the amount of block space and uh use cases or the flexibility of the like L1 uh landscape L1 layer wasn't enough we expanded to rollups which uh have like currently are not fully decentralized uh as compared to the l1s I mean some l1s are also not fully decentralized but uh the the the usual ones we know as Bitcoin ethereum are are or are at least at a sufficient level of decentralization so we have a problem where rollups which we depend more and more on for specific use cases and for reaching Mass adoption because of lower costs and uh heighten throughput for transactions have a centralized sequencers for example these SE sequencers are a part of the stack enable that Define how blocks are included and how these batches are then uh posted to the L1 and these sequences mostly for now are operated by the teams which develop the l2s and so this is only one of one of the use cases where we see that we need uh some way and this also I need to point out these in sequences cannot live on chain or on the L1 blockchain but they they are operated beside the chain in offchain manner or in a manner where the other network has their own valid validat or something like that so there we come to ig lay or other generalized reaking platforms where they enable the AVS services so example a sequencer could be an AVS servers where it rents out cryptoeconomic security so the same one that secured the L1 blockchain let's say in this case ethereum for the consensus and it can also like secure this sequencer and in case of malicious behavior let's say malicious transaction ordering or something like that they can be slashed and so we have like a higher level of security or like a higher level of trust uh in the these uh additional Services which depend not only on onchain but also on offchain uh like operations so so this is the same graph more more or less the same diagram where we see that we went from the security consens execution and application uh stack on bitcoin where okay in the case of Bitcoin the applications were scarce but when we came to or we could say that in the case of Bitcoin the the peer-to-peer cash electronic system is the application but uh when it comes to evm and other evm environments or other smart contract enabled environments we see that we have multiple apps uh and on the app player multiple multiple functionalities and then when we come to rops you have other execution environments so this was the case with the l1's l2s and now l3s or whatever comes next and with the reaking Paradigm we want to like decentralize it further or like diversify it further where we have like uh not only the consensus from the L1 or maybe additional Lance or additional like networks we also have these verifiable Services uh then the execution layer we have like sequencers on other apps for let's say data availability or something like like that and all this to enable like apps that can uh like call upon a diverse set of consensus execution and app layers so scr to so as we said simplifies the process of building trust networks for complex systems enables the creation of secure and efficient blockchain Based Services so as you mentioned these avss stand for actively validated services so distributed systems sling in this case I wrote a layer but it can be also carak symbiotic or any other restake platform and these represent like a new class that new new enabled class of blockchain Based Services uh and as I said inherit various forms of trust uh not only from iium robust network but in the case of igen layer they are pooling from the stake ether pool so the the same pool of ether that is securing the Ean consensus so we have three types of PR programmable trust the first is economic trust so the validators make commitments backed by Financial Stakes that is the most common one we encounter today and it shes and accountability and incentivizes Hest Behavior because in in the case of uh like proven or verified malicious behavior the Staker or the reaker in this case will be slashed for his collateral and which disincentivizes this this Behavior so it is critical for services requiring strong economic guarantees so that would be something like oracles or Bridges where other Financial applications or non-financial applications depend on it so where they're not self-confined so then we have decentralized trust Network operated by independent geographically isolated operators enhances security by reducing the risk of collusion or single points of failure essential for services re requiring high levels of s sensorship resistance and reliability so this is also the case for L1 networks but also for some networks like uh uh Oracle operators where we need like a diverse set of uh dispersed uh dispersed operators or services and the the ethereum inclusion trust so the the guarantee guarantee is that ethereum valid will construct and include blocks as promised Le lever address the security and reliability of the ethereum network so this is when we're speaking about a system that uses the stake deter as mentioned so uh we'll start with ar layer so it's the biggest player in this raking space and it's really moving the goal poost for how how this space looks and uh we'll mostly talk about it but when we're talking about IG L what it is what what it's possible to do with it we also like talking about other reaking providers so the purpose to expand boundaries of decentralized trust in blockchain systems as we said with as well not only onchain but offchain compute uh Noel Concepts we'll talk about them one of the key components as we said is reaking so it allows in the case of igen layer concretely to allow valid validators who stake The Ether or have a liquid stake ether representation uh on in an evm environment or eat main it in this case to secure multiple services so we'll talk what these Services can be but they can be more more than one one at a time so and sign significantly enhances the capital efficiency of St asset so uh it means that the the asset that is used for securing the ethereum Network can also be used for other uh offchain Val securing other offchain operations or be used in other other protocols so not only can we use now the liquid state representation we can also use the liquid restate representation or just use the uh the stake theer in other operations or in other uh in other use cases so uh the token maybe not the most important part but it also uh like U directs the way the igen protocol is moving because it wants to like move away from only having fees as something that that the operators can be compensated in so it expands the also the scope of government Beyond like traditional onchain mechanisms okay that that's pretty common in the defi space but uh also introduces a new new layer of community dur decision making also uh enables Rewards or programmatic rewards which can be like on the protocol level or when uh once the AV let's say Services have uh fulfilled some fulfill fulfilled some CR criteria they can also like stream these to other operators uh to their operators as they think it's appropriate so the evolution of trust and blockchain as we said in the diagram before we had the proof of work then we had like the general purpose computation Layer Two Solutions and I other generalized weaking platforms weaking concept so it allows validators to use their stake T to secure m avss uh okay implementation uh we'll cover that later so uh one of the interesting things that is new to ig is this concept of of inter subjectivity and this is where the igen token will come in play so it refers to truths or situations so first of all what inter what is inter subjectivity even so it refers to truths or situations that cannot be definitely proven on chain so this is especially important when we are talking about offchain operations or offchain calculations which are then submitted to an onchain environment represents social truths that reasonable actors within the ecosystem can agree upon and addresses the limitation of traditional blockchain systems in handling complex real world scenarios so one of the first use cases we can think of is or Oracle problems let's say we have a prediction market and we need to like Define what was the outcome of an election and uh in this scenario it wouldn't be something that we can like verifiably prove with the calculation but we need to like depend on the social layer of uh arbitration to see what what what what really happened so 's approach in this is to have the igen token specifically designed for this issue and it allows uh the it introduces this mechanism of Community Driven dispute resolution as you mentioned so uh it doesn't uh so in example scenario disputes over over Val dat to behavior or intentions rather than just actions are questions in such cases pure onchain data might not provide a complete picture as we said uh it allows the community to weigh in to vote and uh then uh on a combination of available facts and Collective judgment so this is something that's not yet fully defined by IG layer but it would probably be uh it would probably involve some kind of consensus with additional oracles or an arbitration system so the two token system so not only is there the igen token but also the B token so the igen token is the one we hear about so it is used primarily for non- staking activities such as trading or participation in defi protocols also for incentivization of AVS services and it is not directly exposed to staking risks uh providing lay protection for for General token holders also used for governance and so uh that is the usual stuff we we can like imagine when talking about tokens in defi protocols but the be token is something that's interesting so it is a backing token specifically designed for staking within the IG L system it is like subject to staking rules of risks uh same as uh restake eer would be so including po/ forking uh represents a more active participation mechanism so in case of significant disputes or faults within the system so where a lot of the system depends on a on a u let's say uh dispute where we have as in this example of the election uh we have uh the these B tokens that can be staked for either vote so uh let's say in the case of the American elections if Cala or Trump won and how to let's say uh how to resolve a poly Market uh like result in that in that case uh everybody would stake their tokens for one of these outcomes and in the case of either A or B one of the tokens would be deemed uh um legitimate and would continue functioning functioning as it is while the the other one would be forked or um uh as um in a way to put it would be rendered uh valueless so separates taking risks from General token usage protecting casual usage from governance related risks it is really an open question if that is really useful to have like a token that is not the backed one the backed one is at least like an interesting experiment in like these Fork token designs but yeah there's all more more like incentives for uh introducing a token that is like like not really uh govern governance really slashing related uh said okay so uh types of abss uh Standalone applications uh may eventually seek to operate independently of the IG layer ecosystem so those are usually applications let's say Bridges or oracles which uh don't have their cryptoeconomic security secured or like don't have a system that is fully circular or like finished and so they can opt into these restake services to like rent this cryptoeconomic security for the time being we have like the second uh the second kind the foundational building blocks so these benefit significant significant significantly from the interconnected demand uh they find Value in Sharing Revenue in exchange for eques growth in shared security so that is really useful for Stuff where there's a programmatic incentives or something where the composability of the def ecosystem can come into play so some challenges for avss is balancing the need for share share Revenue with eat or igen uh against potential independent growth so uh one of one of these issues is that are um such a service would have like a value proposition that is uh separate from the growth of the I layer or any other rest platform and it remains to be seen how these e or let's say I incentives that that are streams from streamed from the protocol to the AVS service if they are value add to the ecosystem as a whole and the advantages of iron layer okay access to substantial pre pre-existing security pool with pre-existing security pool they mean the already staked eer that is restak to the layer so a lot of eer uh I don't know the exact numbers but uh it's a lot of eer that is that was sted for securing the proof of stake network is already uh able to be used for any AVS services and it removes this friction part where uh you can op in besid earning the let's call it risk-free uh yield from the staking to also get additional like rewards from these avss of course additional risks apply so some key takeaways here it it expands the scope of the trust in blockchain system of the decentralized uh trust uh enhances Capital efficiency the two token uh model it remains to be seen but could address inter subjective governance issues uh with this forking uh forking mechanism okay so we'll just quickly go over some some of the alternatives on the market so symbiotic symbiotic finan is also shared security layer enables users to participate as well in permission staking the core architecture not is not that different from IG layer but uh they at the AVS level or DSS level as they call it they enable more like prog atic uh like adjustments that can be used for where some of the let's say when we have multic collateral system multiple restak assets where they can be used for which area services to which amount and how much you you are exposed to particular risks so here we see can see the architecture it's very similar to the I layer one where we have like the re stakers which put up a collateral backing and they opt into these AVS services or DSS services in in the case of same iotic and these operators like opt in uh The Operators opt in to the network and they run some infrastructure or any any offchain or onchain service that can be uh challenged for uh slashing also they are eligible for rewards okay point so this is also a diagram of the of the architecture the users they use receive uh Network rewards through these vaults they also deposit them through them so we only have like a different layer of abstraction between the operators and the users in these vaults car Network Universal weaking layer uh also so uh they also have this distributed secure Services dsss uh they uh allow uh also to rakers to delegate their assets so they deposit the one interesting thing with carak is that they have like a higher potential for composability although they are not really live yet or like in a permissioned uh permissioned State uh they operate on ethereum in arbitrum uh as of now and uh one of the interesting thing is things is that it's really uh well already well integrated into some other defi protocol such as Pendle or like uh uh some stable providers where you can like put up additional collateral which is not eat based okay uh this is some data from July of this year where we can see that I is the clear leader in the category but uh also a symbiotic lately has been catching up not not at the pace of IG layer and here I don't know how well you can see it but this is just a composition of The Ether that has been restak in these platforms where we can see that the rep Stak ether uh is the dominant one on on uh all platforms aside from kak which has like a more Diversified portfolio of not only eatb based assets but assets such as usdc uh okay let's skip over this one uh the five categories created by aen layer uh as I said it doesn't only apply to aen layer but they were the first mover so they they they have the bragging rights so we taking let let anyone use Economic Security on demand as we said lrts so those are only liquid restake tokens so they allow the tokens which are not only like Stak and restate but also the the restate representation to then be uh liquid and used in other def protocols as collateral or something like that AVS Services as we mentioned like they have like multiple like use cases we are yet to see which use cases will stick the most and have the most sense but some of the ones as we mentioned are like slash oracles restake rollups uh Bridges uh FH computation fully homomorphic encryption uh some uh trusted execution environment committees uh ZK verifiers and so on uh hyperscale rollups so rollups which can use uh and take advantage of uh data relability which is provided by a reaking platform uh in this case IG da is a pro project that has been developed by I layer and it is the AVS most most restate uh collateral most restate eer eer at the moment and apps with credible commitments are evolving blockchains from platforms for code verifiable commitments to enabling credible commitments for any human agreeable effects so the whole concept of this reaing is to enable not only to verify things that we can already verify on chain we're not only code is law but to expand that to divider ecosystem where we can use it in real life with these U verified or at least uh Slash offchain verifications or or let's say offchain operations which are then submitted onchain for verification okay so these are the five categories as we mentioned uh so so commitments some some use cases as you mentioned okay this is just a graph from April of this year showing the pace of the tvl rising IG layer so we can see here well that IG layer had these deposit caps introduced for the protocol as it had uh like a roll out in phases where first first they enabled like the weaking then they onboarded the first a services and now they are like having around 5 million ether that is restak that's the data from April so it's probably more by now but we can see that it has these cascading jumps in tvl uh so this is a comment on the shared security uh they uh introduced a um an update where not only you can opt into uh multiple AVS services with your reate collateral but you can decide how much of each uh so before each dollar of weake assets can be restake to multiple avss after each dollar can be restake to only one AVS uh and so the question was is reaking limited to only one weake so to two two uh two things to get the pooling effects pool security we total stake imagine 10 billion with all stake your Stak equally cost 100 avss even though each AVS gets 100 million of a unique stake however it is possible to make the attack cost with new stake prohibitive as more than 5 billion is necessary to take over the quarium and bundling someone create can create a meta AVS contract bundling a bunch of avss uh so uh in this way if you opt into you can be slashed by any AVS in the W also allows post Market bundling so in this case they U outsourc this part of the unbundling so uh this part of the bundling so they are not like really liable for they only allow you to like stti to one AVS asset but they are open to making it possible to have some kind of aggregator which bundles these uh restak assets for multiple area services so uh this is only like prer token support I just the comment that also moving not not away from ether as the so but moving to enabling any rc20 uh to to be able to to be able for like reaking and slashing also for rewards so this is a part of the verifiable software as a service uh idea or like a paradigm where we think about these Services as extending smart contract boundaries so interaction with offchain services uh where the process is emit emitting of events from Smart contract so the contract signal needs for external data need has a need for external data data computation and it pick picks up Events off chain so external Services monitor blockchain for Relevant events send responses back on chain and callbacks uh results are securely integrated or like committed uh back on the blockchain uh benefits greater flexibility development so enable services like as we said oracles cor processes other offchain compute uh okay this is more for developers but High barrier to entry requires both knowledge in onchain and offchain systems uh developing it the system itself is complex requiring also securing a new network but that is something that these weake platforms are enabling or at least the renting of the cryptoeconomic security so uh the AVS landscape so there are five C categories of avss but also much more added but some of these are like rollup services and for enhancing and scaling ethereum capabilities for Applied cryptog cryptography as I mentioned with ZK and FHA uh General decentralized networks so such as for oracles M minor extractable value management and also for AI inference that's probably going to be a big uh use case in the future because it can like verifiably showcase how uh AIS behave in an offchain environment let's say we can like commit and see and test uh compare the capabilities of some some AI models so a character Wallops co-processors cryptography proofs and others some other also diagram so uh the focus of this presentation was to like introduce first of all what reaking and the I layer or other reaking platforms are and like to show case what we can do for it in defi so in they call it intelligent defi but we can like use it use the term like defi that can like make use of offchain computation of verifiable offchain computation so real time adaptability okay Power by layer AVS is okay so the first use case is like a VP tier volume based Fe Fe tiers so the challenge is implementing volume based fees on chain uh the challenge is that it is expensive so our sourcing computation to co-processors such as the lrange and brevis which are working on it co-processors index and store blockchain data mm's call co- processes to calculate Traders fees and then they return the verified results with ZK proofs it also enables access to be comp competitive and have a tiered fee structure uh use case two Dynamic asymmetric fees for amm so loss versus rebalancing which affects the LP's profitability so the solutions can be like the Dynamic fees asymmetric fees or implementation of uh Z ktls so web proofs for secure real-time external data so kind of Oracle and benefits which like the benefits are like improving LP profitability so looking at it from Another Side as opposed to the uh Dex user uh M redistribution action so auction of the right to be the first up to the pool uh the centralized set of operators wants a leaderless auction uh recaptures most of the Los versus rebalancing redistributes or distributes the proceeds to LPS and in in theory it should enable tile spreads Advanced margin system uh so the current limitation is the inability to Express Cross asset viewpoints so the solutions would be like leveraging cor processors for that something that could like Leverage ver verifiable ofin computation uh so in this case account for multiple factors Catal positions pnls and so adjust the margins consider real time liquidity and open interest and to enable like more complex or flexible strategies uh pricing and option is amm uh so the challenge here is to Accurate accurately price price and capital efficiency for options amms uh solution is to integrate offchain data and compute so for example to calculate something like IV uh which is the implied volatility in options option options contracts so include volatility historical pricing real-time Market signals uh also can take a use of co-processor and as we mentioned this IV let the implied volatility is used in formulas such as black black schols which is used a lot in the options markets for pricing options uh AI system for parameters so this one is like really in the AI agent space but it's also as most AI agents are not really fully on chained they require these uh offchain or verifiable offchain proofs uh or commitments so in this case we can like Outsource the uh lending or let's say the risk parameter decision makers to some AIS which are then uh so they use a f cor processors and the process they can process a vast amount of data which is really prohibitive on chain and then simulate scenarios predict worst case outcomes the benefits would be real-time responsivness to market conditions so like to be more more quick and to like U disable or like prohibit like uh um let's say toxic flow personalized loans offer customized loan loans based on bers risk profile so something like a personalized credit score which needs offchain data or compute so self-explanatory so here we can like also come to under collateralized Landing or based on reputation or something like that uh some additional use case comp compliant privacy mixers uh something that like to cash but uh also while being being able to comply to regulations uh automated rebalancing yield protocols hyper targeted incentive programs and so on also some interesting use cases are with unisoft V4 which introduce these hooks so uh as we as you know the the IG can add like some some added functionality to these hooks so they can run matching engines B transactions run actions access real world information which is really really useful in this case and so uh solve Coincidence of WS to reduce taker fees Implement Dynamic fees for high volatility periods automatically adjust LP positions uh create dark pools for example uh private orders uh run auctions to reduce lvr and maintain shared State uh also one interesting use case is the use of H FHA co-processors which can enable some of these things such as uh these uh where was it uh dark pools so like these uh dark pools for uh swapping uh I won't go into detail here but Phoenix is uh like leveraging I three staking mechanisms so first of all what are FH co-processors they are companion processors offloading specific computational tasks from a host chain so they specialize in maintaining data conf confidentiality during computation so we can have like private data that is still stored on chain but is not really visible to any anybody aside from the co-processor as coprocessors I so I should also add that fully homomorphic encryption enables like operations on data that is fully encrypted uh and uh like doing operations on it without ever decrypting the data so keeping keeping keeping it encrypted and as I said preserves confidentiality and F schemes so as we mentioned like some applications in future impact went over these already and for for the end I'm going to plug plug have a Shameless plug about our project where we uh tried out ourselves at the reaking Paradigm so we built uh let's say let's call it an igen Bitcoin or gazelle Bitcoin so it is a bridg Bitcoin to ethereum to a bridge from Bitcoin to ethereum to bridge the native Bitcoin from the Bitcoin blockchain to the ethereum blockchain we utilize I layer to secure it uh we took the idea or in the inspiration for zerx not going to make it and here's just a quick overview of the architecture so in this case we have Diagon lay core contracts and where the stakers or sters would put up their collateral to which would be uh like susceptible to slashing or rewards we have two AVS Services which can also be one AVS services but in this case in this diagram we have one that covers the Bitcoin to Ether transactions and the E to bitcoin transactions as we know bit Bitcoin is not in uh like a fully cheing complete environment so we don't can't have like any uh we can have uh some uh let's call them smart contracts or some programs but we don't we can't have any computation on it so in this case we implement a Bitcoin multisig on the Bitcoin side so it is like developed in the classic like Bitcoin script language and when the user let's say wants to dep uh deposit when he wants to brige his Bitcoin from the Bitcoin blockchain to have an erc20 representation on The Ether side he deposits to the Bitcoin multisig which is a simple like multisig wallet which is operated by The Operators of our AVS Service uh once see deposits the the AVS that secures the Bitcoin to eat uh like flow uh checks as a as a kind of offchain or Oracle if the transaction really happened and credits the user with this in this case gazel Bitcoin ec20 representation in any evm environment in this case and L L1 or main at eum uh the operators use the Bitcoin light nodes to check to check the state of the Bitcoin uh Bitcoin blockchain and in the case that he wants to redeem his usc2 representation representation he wants to go back to um to bitcoin to the native chain he just burns his ear see2 any representation to our contract smart contract and then the operators like check that burn happened and they uh they are tasked with uh like uh emitting or like releasing the Bitcoin from the multis back to the user on the Bitcoin Bitcoin blockchain this scheme can work because the operators are like um susceptible to slashing uh in in ether so that's one of the considerations with this system that the colal which secures the Bitcoin is actually eer in this case uh it could be in the future some other year C20 or like ec20 Bitcoin representations so uh they are disincentivized from like like stealing the money from the multis and also they need to reach a consensus when reading the state of Bitcoin and also signing the transactions on the mul and yeah that's that was a Shameless plugin you can like follow our team on Twitter uh that's me I'm coota and also TG if you want to reach out thank you any questions I have one mhm uh so when you were talking about this B token you mentioned it could be used to uh resolve markets for like prediction markets like binary markets like markets uh how is this any different than than the Yuma protocol they are currently using uh it's actually similar to the Y protocol it's just repeat the question to the mic yeah so uh this person asked in this case with the IG and the B token and the for King where the dispute resolutions are resolved with staking to either the A or B scenario in this case let's call it the binary scenario where the those who lost or where consensus the option where consensus was not reached a slashed uh if that is similar to the Uma protocol Oracle uh and yeah it is similar I'm not really uh familiar in depth with dor protocol but it is a similar mechanism they employing anyone else d ask a question did you win any he with that project yeah yeah yeah we won the the youth Belgrade heon first place with this with this project the gazelle inspired by the brid over there I hear that's a really cool hackup yeah it is here should go again probably yeah we also invite anybody to join the hackaton next year or it's really a fun experience yeah yes thank you uh any more questions no all right feel free to stay around keep up the discussions about happy I L with some beers thank you for coming and see you next time