Indexing GraphQL Querying System with ZK Proofs

Transcript

[Music] foreign [Music] okay uh sorry about the delay there uh technical details but it happens with my class all the time um so what I'm talking about today this is actually my third time speaking at ECC um but this is the first time I'm actually talking about something that I've been researching for at New York University so I'm talking about smart contract indexing and ZK SQL um I am a CEO of a hedge fund and we're a

spin-off of Ledger Prime we've been market makers uh for many ethereum based protocols uh but aside from that I actually also teach at NYU uh attendant School of Engineering and I've been doing that for like past three years or so but this is the first time um these past six months where I've had a lot of students that want to do Ms thesis that wanted to guided studies so we've started publishing a lot of research uh and putting a lot of research out and this is all academically backed by NYU prior to that I was at Republic and prior to that I was a very early employee at consensus so it's always an honor coming back to HCC

uh this is the published research that actually just went out a couple of months ago if anyone is interested in reading it happy to you know send you the link but this is the first time we put out all the zero knowledge proof of originations and the use cases and for this paper we actually took each and every you know project that's saying you know we're building on ZK and you know whatever they're building on and we tested it out because we wanted to put like you know the most accurate information out and this was published on May 18th it has started to become a

bit of you know background for NYU but one thing I've also started looking with another student is uh ZK sqls and ZK indexing so and if it gets a bit more technical keep it in mind that we're still defining ZK indexing as we go on so ZK sqls you know valid for computation of curing right so serving parse data from databases and indexes is parsing data from blockchain which currently only supports fixed security SQL coming from you know the originations of you know web it's a language for curing the data from a

database uh how we add ZK SQL it's providing validity for curing or inserting operations but this also secures that resulting data and database itself is valid um ZK indexing is a unique indexing infrastructure that excels in both accuracy and security using zero knowledge proofs and it's a typical output through ZK Oracle with data flowing from on chain which you know in ethereum in this case to off chain we've been testing uh graph sqls so you know in this case we're doing uh using ZK graph index data

um so far we've seen ZK SQL with option databases so proof of SQL is one example uh they're building a data warehouse for blockchain data and they're making sure that the data is indeed from Warehouse uh that actually also means that the data does not need to be 100 the same as on chain uh why does it matter because when we were when I was at a hedge you know when like Ledger Prime was up and running uh on-chain data was something that we struggled with through and through and especially you know in ethereum's case uh because we wanted real onshine on-chain data to make the real-time

trade to real-time make Market Market make on it so could this be a solution I mean hopefully I would have a better answer by like October November and then ZK sequels um we've started seeing like some good luck with around circum so zika SQL proverb nodes maintain the database and generate proofs for each SQL query um if it's on chain database Zeke is equal with on-chain database it's the same for individual smart contracts data or whole blockchain state it's like literally an on-chain oracle with SQL as you know the query language so some early exploration

that we're testing on chain SQL without ZK compound lens and squalidity uh I you know six months have passed I still can't say the word uh but it's an on Shades Clyde for smart contracts the gas fees you know we've started to have an issue with it um it may be too high for maintaining this on-chain data the fact that we're testing it out you know quote unquote in a bear Market ethereum isn't as expensive as it as it was in 2021 we still realized you know for an everyday use it's going to be a bit more costly so

um you know adding ZK to it you know without off chain ZK proving to remove the on-chin computation and then tableland is very much like the graph but uh you know database for web three depths um I don't we they haven't published out a lot of research on them but uh for for from the basics of it uh they're following a very like similar model to what you know graph sequels is following um ZK is equal for on-chain data where the open market is that's where you know our research is so much focused on so no current ZK SQL for querying on chain data uh but graph protocol is working on ZK graph graphql curing which is very

similar to ZK SQL for onten data but sellers brevis is also working on ZK plus ZK graphql now keep it in mind um we have started researching on this uh like late last October and we've covered up the period up until like early July and the market or you know the tech stack or the interest levels have changed it quite a bit in the ZK frame uh but for this current research we're going to follow ZK graphqls uh you know thoroughly because the graph you know as

a project itself you know have been phenomenal with the ecosystem so anyways yes so ZK sequels need to implement all operations of SQL including select insert and delete ZK SQL for on-chain data makes more sense because it can show the advantages of ZK including small verifications cost benefit costs better than DK SQL for off-chain databases and I think that's it because I got my five minute Q also and I told you I'll be on time uh but the point is that look this is you know if you guys are builders in the room if you're building on DK and if you're

building on you know ZK hardware and whatnot on the backend side of things Zeke is turning out to be a very like promising ethereum solution so there's a lot of academic research that started to go into and that's why you know when we pitched it to NYU you know they said yes you you have a blessings please go ahead with it um would we have uh when the published paper is coming out on this early next year because now these days we've started to integrate ZK hardware and see if that could solve a solution but I'll probably have a lot many more answers early next year on where ZK sequels you know tend to go thank you

foreign do you have questions do we have time hello yeah this is junda from seller Network thank you so much for the Kyoto brevis yes [Laughter] thanks um you know just also let everyone know the brevis we officially kind of make the alpha available and everyone can go and try I think the the core idea I think you know the Unchained data has the ground shoes right on the blockchain uh the grow so that's the part we always

have been struggling with right so the monitoring unchain data save into a centralized database then you have to trust the database and then you know hopefully as the technology the zero knowledge technology keep making progress is faster and cheaper we could remove the the centralized database in the middle right we can go directly from the Unchained the ground shoes with a succinct proof then you know go to wherever the chain you want to use it my question is I actually bought ZK nowadays you know it has been used a lot what's your feeling about the kind of the technology and how does it even represent now right oh God uh please don't like hate me in the on

this but the thing is that the hedge fund me is very like bullish on it but the NYU Professor me is saying oh God this is going to take a while because we've started like I said that we've been we've been researching on it since last year we've already published a paper we're about to do a second publish it's very promising but it'll be a while so that's why we wanted to see where does the head you know Hardware stand so that's that's becoming our next Focus thanks for sharing about a topic I didn't even know about so I have a pretty Noob question I think I understand how ZK proofs work for on-chain data because you end up tying

it back to some block cash or some cryptographic signature or identity but for a centralized off-chain database what is how do you know that the proof is Tethered to like that database or something that you trust without some sort of blockchain like provenance forgetting the name but they're basically using I think proof of time uh to go uh and there's only one project that's doing it it's actually mentioned on my second or third slide I mean we're short on time I can talk to you uh

outside about this but uh it's called proof of SQL actually consensus mechanism is strictly built for like keep the mic on please strict strictly builds for off-chain Hardware now is it proofable is it tested it out on a very small scale we tested it around on 10 transactions at work put 15 000 transactions on it I don't know like honestly speaking I don't know um thank you we still have time for one or two quick questions

if there is any other ways thank you so much [Music] [Music] [Music] thank you