DAPPCON 2024: Intents Are Not Real Without The Intent Engine - Connor Howe

DappCon 2024 · May 2024 · 24:21

Connor Howe argues that intents are not real without an intent engine, at DappCon 2024. In Web3, an intent expresses a user's desired outcome rather than the exact transaction steps needed to reach it, and an intent engine is the system responsible for interpreting and executing that request on-chain.

Speakers

Topics

From DappCon 2024, hosted by Dappcon Berlin.

Transcript

I'm Connor how one of the co-founders of Eno I've been building an ethereum fi time for the last eight years prior to that it was in Bitcoin in 2013 and Enos went through many Evolutions to be able to get to where it is as of today the agenda of today will be giving you a bit of context of the evolution of Eno so you can get some insight into how we ended up to intents what are intents and the Eno Network the Eno network is not being publicized anywhere there's no blogs there's no tweets so you're getting real real Alpha today so crypto projects generally have two choices either artificially create

product Market foot by launching a token hopefully you survive after 2 3 years most don't or you go through the fog of unknown that's what we done with Eno we started off with a vampire attack we attacked six different protocols gamified this whole experienc and incentivized it with nfts we then launched social trading where anybody could create a social strategy of monetizing your investments into Defi and erc20 tokens whil building this we found a massive massive problem was how do we integrate all of dii Simply conveniently without

spending over half a million of dollars in audits which we did only for 15 D5 protocols so went back to the drawing board rebuilt everything from scratch and had one holistic UI for interacting with all of defi whil was building this the architecture underneath we built in a very nice way that we're able to integrate a new D5 protocol from inception to production within 30 minutes safely but users again don't use a product unless you give them free money which doesn't really make sense but we got a light bulb moment which was everyone else has not everyone but many people have the same problem as

us in the V1 of how to integrate all of defi into your product we spoke to many projects and teams and two results came out from this the question was how do we get users one launch a token two we have the same issue so we spun up an API and within the first week of the API we got 11 million volume more cumulated over the two years of Eno being around so that's what you know Eno as of today this def5 middle layer enabling application developers to embed defi actions into their products ranging from lending borrowing swapping LPS involves so if you're a developer and you're

building a wallet a treasury management tool whatever it may be and you want to offer your users defi you don't need to build it all over again manually one by one you can simply just call our API for this and in a matter of 3 years we went from first year of $8 million to 200k to now 13 12 12 billion plus with 35 Plus projects using as in production all without incentivizing this with a token whilst we were building this we always had user centricity at the Forefront users want to interact with dii without needing to know every single

individual complexity of how to interact with that defi protocol so let's say for example you want to interact with Bey that requires an aura LP that requires a balance RP that requires three individual tokens with inso you can just say I want that balance or LP and we'll abstract all that away for you or bfp we called this shortcuts of blockchain other people call this intents for those who don't know what intents are they're really not that complex some people like to make them sound more complex than what they are but I saying I have a and I want b c and d and I don't care how you get that for me

all of this logic of how to get that is abstracted away for the developer and b c and d could be a token State nft defi position or any interaction on shame you may have heard a lot of hype about intents and if a user wants to express their intention to go from token a to a D5 position you'll go to what people call an intent netor networ there's many different intent networks as you can see all these intent networks inherently are is a job bulletin board so you go to LinkedIn jobs you go to Fiverr you go to Glass glass door upwork you say I have a which is money in the context of a job

and I want this task to be done it's the exact same for an intent Network and underneath people that fulfill this task or this intent are called solvers some people call them resolvers some people call them fillers some people call them transmitters whatever name somebody creates next is the same thing I have a task I have a a and I want b c and d and somebody figures out how to do that for you but all sers really are doing as of today is swapping from token a to token

B there's only two players in this whole ecosystem that are doing njust token swaps so for example if you want to be a solver you're creating this map of all smart contract interactions token a goes into this contract token B comes out and you start to graph all of these relationships to give you a visual representation this is one of the orders we fulfilled this is doing 100 12 different individual transactions to be able to obtain the most optimal route for the user so you start to see that these are individual points in a map and how all these relationships start to connect

with each other yeah I mean you can see the extent of this right it's quite big not going to explain everyone but again the ecosystem is not just about tokens b c and d could be a token a D5 position an nft a state outcome could be an approval it could be a transfer it could be a mint it could be any smart contract interaction on chain our vision of this ecosystem is all developer interactions will be

intent based in the the future defining what you want and not how you should get that makes the most logical sense as a developer that wants to offer maybe a DI product a DI protocol in their product they need to integrate maybe four or five different dii protocols before they can even offer that to their users so these developers are turning into a Dev shop an integration house for building every single integration waiting for audit slots looking at four five 6 months before you can release an MVP of maybe just one D5 protocol your user asked

for this is where Eno comes in so we have built the world's first fully decentralized shared network state for building transactable data across all blockchain Frameworks is his own blockchain and this choice of a shared network state to really just break this down the world is a map right you have many different countries in a world you have many different cities in a country and the Eno state is a world of all smart contract interactions across all blockchain Frameworks so in this case ethereum can be represented as a country each individual smart contract you can

see as a city by having all of this within one place we can then start understanding the relationships how they intertwine how they communicate with each other not only for tokens but for any smart contract interaction on chain similar to how a train line may work you want to go from point A to point B you're going through many different cities to be able to get there with this as a full map we can start understanding the relationship ships of smart contract interactions and all of this turns all developer experience into purely outcome

based if you have all smart contracts how they intertwine how to build executable data of being able to only specify your outcome developers don't need to care about all the individual nuances for every single smart contract they integrate developers can now just say I want this inad instead of each individual step they have to build manually so developer would come to the an Network specify their intent the intent could be again an nft defi position a token a state override and all of the individual parts to be able to obtain that is abstracted away with

the inso network for example even approvals it's required as a prerequisite is abstracted the three different network participants we have is an action provider a grapher and a validator we have competition but also collaborativeness between Network participants an action provider you can Envision is somebody coming along and saying here's how you interact with a smart contract conforming to a standard that Eno has built similarly how a developer goes to subgraph well the graph and creates a subgraph that other people could consume let's say for example you want Avy lending you will

come along to the Eno Network here's how to land on Avy this now is a plot inside of the Eno map the graphers will listen to intent requests these are purely mathematicians that do not care about where the data is coming from mathematicians don't care that it's blockchain data they just Traverse every single action on the Eno map for building the most optimal solution these graphers will propose their solution of bite code of how to solve the intent into the mol and the cosmos Jam validators will then simulate this and then find the most optimal solution that was

proposed these validators in the background are running transaction simulation tools we open- sourced our transaction simulation tool about 10 months ago it's called temper so if you are using tenderly you can just use this for free so in this F map of all smart contract interactions you need people to contribute actions to the network so a developer maybe in this room comes along and says I want to add the substraction for the smart contract if you add this you get a fee of all solutions that are being built using that action the

question you may have is what does this turn into domain squatting not necessarily because there's a full economy of token staking on top of the actions that get provided so if an intent solution comes in and two actions are the same the one that has the most amount of inso token to Stak get selected for that solution why tents aren't real without the intent engine intents are purely just job buletin boards you need people underneath to fill these different intents we cannot build hundreds of thousands of drob buletin boards when there's only really a handful of solvers in ecosystem that are only doing token

spops as of today as this ecosystem to ecosystem grows which it is with the amount of blockchains the amount of layer twos rollup as a service we're turning into this real multi-chain future and we keep forgetting the developers want to inter integrate all of this if you integrate one blockchain you have to build months for this have all infrastructure for this and then integrate each smart contract this does not scale integrating every single train manually itself and we need people to use this technology if we expect developers to integrate everything manually it doesn't

work we learned that ourselves the hard way in the version one of Eno spent over a year integrating 15 defi protocols over half a million dollars in audits if we want people to build products on top of this they need a convenient way of embedding all blockchain into their applications so we're to build but whatever however wherever using Eno abstracts all of this way for you to embed all blockchain into your product the only limitation really is your ination now you don't need to spend months integrating every protocol yeah that's a presentation if you want to contact me Conor and in

Anda still got five minutes you know um we are hiring so two Engineers a Twitter intern if you like to ship post let me know developer relations calls most developers and if you think you are oh well maybe shouldn't be hiring you know exactly is it the head oh there you go but if you think you and we does I turn my head not um just reach to me and yeah thank you if anyone's got any questions thanks for the talk um staying in your

map analogy um how do you add new cities to your map is it a Google Maps or an open street map like new smart contracts like staying your map an analogy how do you add new smart contracts to your protocol a standardization of different Primitives you come along and plug it in and you say here's how you land on ay here's how to execute it's what anyone can do it's not really deploying a new smart contract it's just like an interface layer of knowing how to interact with these other smart contracts I assume the proofs depend on a certain state of the blockchain so how

do you make sure that the blockchain does not change in a way that the proofs are not valid when the actual ex execution happens so there's no real proofs inside of this it's more once you cont action to not only I didn't say in the presentation but you specify how to execute and also how to validate that execution happens then the validators are then going through each individual action which used compris of the solution and then also validating insight for these individual ghs and there're also so maybe the getter is not there validator can start adding custom modules for then State

over um i r a bit of infrastructure I can quote how can I participate in answer so there's three different parts if you can quote one you can build smart contract abstractions and just contribute them to the network I recommend you to to start doing it whenever it's live from day one because all the solutions that are being generated you will get portions of that fees so it's kind of like you're front running other people contributing to the network this the second part I wouldn't recommend graphers if you're AER this purely mathematicians and then the third one is a validator so if you've run any nodes before just take this image and then

just run it what about the gra yeah I wouldn't really say it's more I wouldn't think that's a coder it's a purely mathematician so you're doing like K shortage paths You're Building costal go to Traverse a whole map you can do that if you want but I think validators and action providers probably easier for you oh I've got two now okay double bmy uh thanks for the presentation uh maybe just one quick thing uh so if I wanted to say I have a dab and I wanted to offer my users

to deposit something some some funds on a and do something with the yeld uh I had to build an intent and then uh I could use that intent and anyone else could use that same intent um it's a really good question right now no so you're basically saying specify and then storing this somewhere so people can reuse what you've requested right no but what you can do is a dow you can just call the NSO API and it's fully abstracted away so you would store most likely in your back end hey I want to go from A to B and it's could be landing on a so you know that call would generate the call data and then the user would execute themselves if you do want to

make it more funky you can use gelato web 3 functions so you can kind of automate this whole process under these conditions if apy on Avi is X call NSO API get call data and then also gelato can execute on behalf of the smart wallet or the smart account thanks thank you uh I'm just wondering what is your business model how are you charging um so right now it's kind of taking spread comparison to other

people so the assumption is not only are we going to have direct developers using us we're also going to have other intent networks consuming us so for example like we're soling C swap Brink users aperture many different intent networks we can see what our Solutions are compared to other people as well inside of this and then we can take the percentage difference moving forward most likely a fix fee variable based upon how much volume is coming through this whoever's consuming the network right now TBD okay and then one follow on question then you said uh you can see how good your Solutions are compared to other Solutions and that solution you showed was really looking highly complex I'm just wondering what kind of uh

Improvement are you currently seeing your Solutions offer over the best in the market comp to you so we're not the best at token swaps like like probably not public knowledge but now it is um we're more good at defi routing so like we one with another person um are the only ones doing defy routing so our Edge is being able to interact with a new def5 protocol very quickly let's say you want to get into SD many people try and purchase that on a secondary exchange but the most optimal route maybe swap into X different tokens and then mint SDI directly so that is kind of one of our edges okay cool thank you yeah hi so what part of the solution

are centralized what are offchain onchain because if we are talking about competitive Roa roting solutions for example like Unis X or C swap or solvers so they have different type of centralization inside the system you asking about them or the answer no I'm asking about you so as I understand this interfaces where are not on the blockchain where on the back end or on the API levels so what parts are on the backend what parts are on the blue offchain you know so as of today it's fully centralized once the network goes live it will be the validators first and then over time will be individual graphers they'll the graphers

will be running their own proprietary alos offchain they won't be putting that onchain they'll be putting the solutions generated from their alos on chain then the action providers that's put on chain so people saying here's how to interact with a protocol over time it'll be individual RPC requests that are being made to the network itself M okay thank you uh how far do you abstract away the intent so for example if I want to take

on let's say of a $1,000 in ethereum and I want to take on 1.1 leverage on that ethereum and like I I do that through Landing markets like do I need to specify in my intent that I want to do this on AA or maker or can you can you kind of abstract that away completely yeah so I think people will build tools like that on top and it will turn into three different types of intents explicit I.E I want to do my leverage on that protocol that's like explicit semiexplicit is I want to obtain this leverage but you're creating these boundaries that solvers can then still use so maybe it could be an a maybe it could be maker maybe it could

be GMX whatever it may be and then the third one is not really explicit at all and it would be a AI execution B this learned from your previous Behavior what you would do in these market conditions I want more money you tent really though really thank you