Decentralized Infrastructure Network (DIN) as an Eigenlayer AVS | Iris Radoi & Nate Fikru -Consensys

welcome back so uh just a quick announcement so at 6:20 there's going to be a screening of uh vitalic and ethereum story at uh the captain ethereum stage just behind me so be sure to check that out uh next up we have Iris rudoy and Nate fikru from din of consensus with uh decentralized infrastructure Network din uh Journey as an Igan layer AVS hey everybody all right so my name is Nate Fu we have Iris right here uh she's getting her mic connected I got you okay okay AR can everybody hear me okay sweet thank you all right cool so good evening ladies and gentlemen I'm Iris as Nate said and um I a product owner at infura working on our decentralized protocol decentralized infrastructure Network or for short DIN and besides me I've got Nate who is our lead engineer and together we're going to dive into our journey uh within as an igon layer AVS and really and truly how we're involving INF fur's Legacy into building a truly uh decentralized web free infrastructure Marketplace so today we'll kick things off with uh a few din Essentials what it is why it matters and who is behind it uh and then we're going to move into some Marketplace Basics and then I'm going to hand it over to Nate who is going to dive a bit deeper into the crooks and nitty bitty details of Dena and abs and then we're going to move on looking Beyond uh our next few Milestones that we've got in place all right so first things first what is din it is a permissionless Marketplace for web free infrastructure where we connect providers with gateways uh for delivering rpcs apis blockchain data to developers across a variety of blockchains since if Denver last year where we started off with offering support for one service we've made a tremendous amount of progress and I'm happy to say that now a year later we do offer support for 30 networks uh and we're also serving over 100 million daily requests on our Network and um this wouldn't have been possible without the support and the enthusiasm of the entire RPC community so we're very thankful for the support we've received from the providers who have partnered with us so currently we have 50 Partners on boarded onto the marketplace and 15 uh providers who are actively serving traffic AKA The 100 million request I just mentioned earlier so this is really done by Design because we're really trying to build this in a sustainable way where we are ensuring top performance for the users as well as economic viability for the providers so we are working on on boarding more and more providers as it makes sense and and we are doing this in a fashion in which we can ensure that the network stays reliable it stay it's scalable and it's also providing the quality of service and the great user experience that we're building towards and um with this I just want to say that D is not just building for scaling but we're really taking our time and we're building uh a de centralized backbone with purpose so now I just want to dive a bit deeper into why we started doing this in the first place so I'm just going to throw a few stats at you quickly so in the beginning of 2024 cloud service provider outages spiked to about 27% so if you really think about it what that means is in some way shape of form some companies lost a lot of money and there was some really bad user experience for a period of time also at the same time we have about 63% of devs that cited that uh reliable infrastructure is the top bottom leg in their development process so with these things in mind um we can kind of sort of see the pattern as to why we're starting to build build in the RPC space is blowing up there's new chains popping left and right and everyone's trying to get in to be the first one to serve the chain to get on the trend and this is cool but as a single team building RPC Services you you face a lot of challenges because number one you need to make sure that you provide top service quality number two you need to make sure that you're economically viable and sustainable yourself and you're also staying competitive across the hundreds of chains that like are appearing on the market so really and truly we are finding ourselves in the famous saying that people are looking for the Jack of all trades but as we all know this is followed by being the master of none and with this in mind uh this is really where then comes in because we what we're really doing is we're offering RPC providers and single teams the opportunity to really focus on what they're good at and then being able to join a marketplace where they can shine through and gateways can tap into these providers who are specialized in their specific area of expertise and they can quickly on board chains to offer that sort of user experience where you get the Jack of all trades and the master of all trades so uh now I just want to quickly move into to uh who is behind in and who are the actors in our Network so I've mentioned a few times already uh in this presentation the term gateways so these are entities that buy resources from the network and they sell it to their customers so our first turn boarded Gateway is in fura who by using din is providing access to its customers to chains such as Stark net blast Mento and uni chain just to name a few next we've got the note providers these are really the building blocks of the network because they are the ones running the services that everyone is going to buy and uh the great thing is that being a Gateway and being not a not provider are not mutually exclusive so you can run the services that you're good at you can put them up in DIN and then you can also tap into the marketplace yourself to on board chains and services for which you don't really want to deal with the problems and the uh headaches of running this in-house and last but not least we have the Watchers which is a key component into our uh ecosystem because they are the ones who are publishing Insight of reports about the performance of node provider providers and they're really ensuring that the quality of service is there from the moment where we on board the provider where we run benchmarks to make sure that they are up to par for the Stander for the network and also continuously monitoring the performance of the network as it is live cool so before I'm going to hand over to Nate to go into a bit more technical details I just want to do a few quick Marketplace Basics with everyone so the easiest way to think about this is just to take a real world example so let's just think about the farmers market what do you have to do in order to make this successful and to make sure that people show up so first and foremost you do need to have a location where the market is going to take place next you do need to have suppliers AKA in our case local Growers who are going to come to the market to sell their goods and the reason why they do this is because they don't have to go out and individually look out for customer to sell to because they know that by going to the market they're going to have customers coming to them uh then um once they come to the market usually what happens is they would have to pay a fee in order to get access to a store in the market where they can sell their goods and then once the market starts you normally have customers who kind of just browse around and they mix and match and find you know one thing from one provider that they like another one from a different provider so in a nutsh show if we abstract this sort of real world example and we take it into our world into the din World things work kind of very similar so in our case the market place is din itself so this is the location where everyone kind of gathers to sell their goods and you have a series of Watchers and gateways who are getting onboarded onto the marketplace in order to sell their services and normally what happens is they would pay a an amount of capital in order to take part in the network to ensure that they you know have the sort of stall to be able to Showcase their services that they are uh have available and then um once we have once the market is open you would have customers coming in and then they would pick can mix what they like for me whether that's like your price your speed your performance and it's a really a mix of all of these things and then the re real difference between the farmers market and the din Marketplace is that din is going to be running 24/7 you can always tap in you can tap out things are always going to be updating and with this I'm going to hand over to Nate who's going to go into a few more details about Dena and AVS thank you Iris thank you so much for that amazing overview of d uh and what we do every day so my job is to pretty much tell you guys what we're trying to do with Den as an AVS and why it makes sense so to begin off Den started in a Federated state so that's pretty much the idea of uh a small group of really trusted operators running the traffic through our system so that was the intention of D though was to always be moving towards a continuously decentralized model so we wanted to take steps from a feder ated State into uh you know more and more decentralized as we get bigger and we uh increase our experience within the space doing what we're doing um so with that being said what we want to do is utilize igen lay to actually uh plug into ethereum security um the the security that ethereum actually gives you at the the bottom layer and uh we do this via reaking I'm sure you guys have have heard about what reaking is so as a result we actually don't need to run our own validators D we can we can piggyback off of igen layers protocol and what they already have set up and this also aligns with us um when it comes to staking because with igen layer you know because staking is built in we have a a framework for essentially putting in rewards slashing and that sort of thing depending on service so to talk about like what the components of an AVS are uh especially with D um you know we have operator sets these are kind of um any entity within our Network that's going to be staking so these are most likely going to be our providers at least at the in our beginning phase and they're essentially going to be uh offering services so that you know for example they could potentially be offering uh polygon Services uh within din so that would essentially be uh considered an operator set next you have a quorum so this is actually a subset of operators that are running um within the uh within D and we different differentiate them by essentially what uh rule set they're running under so they might be running uh essentially staking for specific token for example like a specific restake eth or maybe another type of token so that's what a quorum is um with this entire system we have rewards and we also have slashing so with rewards we incorporate the ability to reward our operators based on the performance that they're actually giving to the network so if they're giving good performance and they're not doing anything you know wrong and you know they're good actors they have the ability to actually gain you know rewards in the form of a token of uh their choosing for being good actors on the actual protocol and then lastly we have the concept of slashing so like uh opposite of rewards you know we you uh we actually are incorporating slashing multiple different uh types and I'll explain that in a second but in a uh in a nutshell it's essentially a penalty for misbehavior so staking and slashing on D so this is kind of a core concept of why we wanted to move uh D onto igen layers so they make it really easy to stake like I mentioned and to uh get you know any sort of uh token pairing that you want as a Staker um staking can also be tied to the actual performance uh the amount of traffic that uh an operator would actually want to receive so we are actually going to cap that off at 20% of traffic by default based on how much uh stake they're contributing because we don't want anybody on the network to be too overpowered but we do want to incentivize people to incentivize uh operators to stake more so like I mentioned there's kind of two different ways that slashing happens so we have the concept of a minor slashing event this is kind of like uh an operating return returning back like a 4 400 response maybe uh longer latency times that sort of thing these AR major events to the point where we want to um we just want to reduce the amount of traffic that actually goes to that specific operator so it can act as a slashing event but it's not actually slashing in the traditional sense of the term whereas a major slashing event this is more of the traditional um the operator does something like extremely wrong on the network for example returning data of a different network or a different blockchain and this is where their stake actually gets slashed so I think we're we're thinking around the the 3% slashing rate but that can vary once we're live excuse me and lastly we have the concept of a veto committee that comes out of the box with Igan lir uh our operators can appeal any sort of slashing event if they think that something wrong has happened and um we have a group of top operators within the network that are going to review that entire system so we've been building d as an ABS for around the last 6 months um starting from the feder state and slowly building it out as a decentralized format um we were lucky enough to be able to partner with alt layer and igen layer to help us architect the system we work with them directly and uh very closely and um they've been able to help us integrate the actual igen uh middle middleware and contracts that are in place um and this work has been going on ongoing for about the last 3 months um we've been able to build Uh custom logic specifically for Den regarding the staking onboarding verification Etc and lastly um we're still in the process of building but once we're uh a little bit further down the road we'll be going through an auditing uh phase where we'll audit some of the smart contracts with in our consensus diligence team as well as the igen lay team will be auditing and of course we'll most likely have uh other third part party Auditors going through our contracts as well all right so looking Beyond as you see we've got a very interesting looking road map where we don't really have exact timelines of the Milestones that we're trying to hit I think as we've seen in the past week things are very unpredictable in the ecosystem that we're working test Nets where you're developing can always break and you don't always have control over this so uh we're very close to launching our first AVS fmal test net and we are as I mentioned earlier we're doing this in a way where we're very calculated and we want to make sure that we do things right so we have at least three fmrl test Nets that we're looking to launch in a phased approach and from each release we're going to take learnings and we're going to test uh different criteria we also will soon be launching a beta program to add new gateways onto uh onto Den right now our main gateway is in fura and uh that's our Alpha Gateway but eventually we'll be growing our uh Gateway list very soon and again all this things are happening in parallel so you can imagine as we're working on the beta program for the gateways we're launching like our femoral test Nets and also we're going to reach a point where we will have a mainnet launch for AVS which is going to really bring us closer to the decentralized experience that we want to offer providers and we'll also be uh we'll be starting off with a command line interface essentially to integrate with d but we'll eventually want to have an actual application a web app that uh operators can use to onboard verify them eles as well as uh Network participants can kind of see the overall health of Di and last but not least the cherry on top obviously we're relentlessly working on building a dim Foundation as this is very aligned with our ethos to decentralize uh the network and again this is a milestone that we're working towards building as soon as possible and while all of this is going on we're constantly adding new networks and new services to what we offer at dim um like Iris said earlier we offer for about 30 different Services right now but that number is just going to keep going up hey in a year who knows it's going to be 100 plus who knows so um last but not least I we do want to thank all of our providers and our partners who have really helped us bring this Vision to life specifically uh we've been working very closely with compare nodes with the rivet team with alt layer with igen and with all our RPC providers who have joined the network uh from the beginning of our journey and just a few technicalities did throw some numbers earlier they're not out of thin air so I do have a few articles if you want to dive deeper into kind of what I said in regards you know the class service provider spikes and stuff like that and last but not least if you do follow along on our journey we try to make this as easy as we can for you so give us a scan give us a follow give us a like subscribe all of that jazz and thank you so much for the time and thank you everybody for being here and we wish you a happy conference have a good rest of your weekend [Applause]