Unmarshal is building a decentralized network for blockchain data

InterviewNovember 22, 202426:45

In this episode

Ashton speaks with Manohara K from UmMarshal on Blockchain Data Indexing, Building a decentralized network for Unmarshal, their incentivized testnet, and how to get further involved.

Key takeaways
  • Unmarshal is a blockchain data indexing protocol that has indexed over 60 chains and serves approximately 7,000 clients with data infrastructure.
  • The platform provides APIs, notifications, explorers, and no-code smart contract indexes to reduce development time and costs for blockchain applications.
  • Unmarshal is integrating AI and natural language processing to allow non-technical users to query blockchain data without SQL or programming knowledge.
  • The next phase involves decentralizing the indexing network, allowing Unmarshal token holders to become indexers and serve data to applications.
  • AI agents will enable users to filter tokens by multiple parameters like liquidity, trading volume, and market cap using plain English queries.

Chapters

Transcript

Read the full transcript 4,162 words, auto-generated

I'm Ashton Addison from the crypto Coin Show and today on blockchain interviews with manahar founder and CEO of UNM Marshall welcome to the show and thank you so much for taking the time yeah thanks for having me here and then it's great to meet folks here yeah I'm super excited to dive into Ai and blockchain uh the intersection of the two uh where we see the future of AI which is is

growing so rapidly uh and how blockchain benefit from that and vice versa I'd love to start off Manara with what you and your team have built at UNM Marshall and how you're helping this perfect match of AI and blockchain and then we can dive into all the latest details yeah thanks thanks ason for having me here and uh uh that's a great first question I think uh I I can't uh

explain enough of what UNM Marshal means to us and like what we've built so far right so UNM Marshal is a block data indexing protocol like we are one of the infrastructure players in the web 3 so we I think like the idea all along was to build the foundation better so that others can build a better D5 better nft and better uh a gamify project so we

kind of serve as a middleware between the blockchains and people who are building on top of blockchains by helping them uh have a better access to the data um and and then we we've done a decent work in last 3 three and a half years we've indexed more than 60 chains uh both evm and non evm and we have helped close to 77,000 plus daps uh we are a B2B company

so we've helped more than 7,000 clients to access onchain data and build their apps faster um so this is some of the I mean like we start more than 10 billion requests uh and like built more than five product lines uh all the way from apis notifications explorers uh and then no code smart contract uh indexes Etc so that's the track record uh and so far

and we are super excited for the next phase of our uh uh our journey which is uh decentralizing on Marshall stag very cool um and I would love to dive into that in in a little bit here about uh the decentralization of it which I think is important I would love to focus a little bit more on uh the indexing of data why that's important for the end users that are trading or

investing um and using other use case applications uh so one a great example I'm guessing is just the price data feed um for is obviously an important part of data can you explain exactly how on Marshall is making uh block blockchain data indexing better for the end user that for example is using a defi application yeah so UNM Marshal it doesn't come to uh come to the picture

when when when end user is accessing some some app right like it may be D5 protocol or it may be simple Investments or it may be accessing his portfolio in a wallet right we sit behind those applications that you use for example if you have used wallet it may be cyp wallet it may be coomi it may may be Frontier it may be uh it may be other wallets right so the the indexer is is a

one step behind those wallets so we Ser the data apas all the price feeds all the uh balances wallet balances transaction history your portfolio within various dii applications like a compound Unis swap LPS Etc so how much liquidity provisioned that uh so those are the information that we access and we build and serve it to your wallet I think without indexer I think your your

apps favorite apps might take uh twice the time if not more uh to come to the life but with with infrastructure applications like us uh the development time of such applications reduces and the cost to build such applications also reduces uh and uh I I think like the old decentral ization is to come out of that block box uh and behind the screens uh

and then it's it's a way of bringing indexing to the common people right um so if you are a marsh holder you can become an an indexer so we build all the tech all you have to do is run that index or serve it to whoever it may be it may be uh the apps that come from unmarshal or it may be direct applications who wants to access the data and so far what we have done is uh

B2B segment like where others use that data and then like other applications uh use UNM martial data and then they build useful applications but with the help of eii we are making sure that even the uh no users oromi users who don't even have to have app can have direct access to the index data where they can go to a app like xcan or they can go to app like

Saki and then like ask their questions in plain English uh without having the knowledge of uh sqls without having the knowledge of quiring the data from databases uh you can directly go and ask uh questions in English and uh let on maral answer this that's the uh one of the major uh features of AI that that uh these this I think like the Char GPT has

brought us right so the way of talking to the systems in your own language uh using llms and using chart gpds of the world so that's what that's the experience we want to bring to the blockchain data uh and and we are excited to bring that forward we've already done that in a form of Saki which is for product managers and uh founders of the companies who want to

access their companies data and what we are going to do next is Agent is each dab so that the community of that dab can also use uh data query data using their own natural language so uh that's that's what uh excites us more so we are building continuous innovation to make sure that uh everyone in the web3 space irrespective of whether they

Builders their investors have uh un interrupted access to the blockchain data to make the additions uh better for their own purpose it may be investing it may be building or it may be any other reason MH yeah I love that because there's so many Avid fans of web 3 that want to also utilize AI but they're not programmers and they want to use these AI agents and llms uh in a way that

maybe they can get the benefits of what a programmer would be able to do without having to know a language and I feel like AI has made a huge leap in the last two years where people can almost code anything and with the help of the AI and they don't even need to know programming language yeah exactly and like one of the use case that I want to give you is

uh in without AI if you want to filter uh give me all the uh Solana based tokens that are listed on binance and have uh liquidity more than 1 million and then have traded uh volume worth 10 million and have the market cap of 100 million right so asking such a question and like accessing that data without AI is is nearly impossible for uh Nami users right but uh but but uh like with

the AI it it becomes super simple and all you have to do is just say what you want uh to a AI prompt so that it can get it and with the intents um you can also do something like okay put my $500 in in this bucket and like have a stop loss of uh 10% plus a minus 10% right so uh that's that's that's going to be another major Milestone that that

uh people can build you leveraging Ai and we want to make all that possible using UNM Marshal data either through directly or through with the partners uh so that's why you see UNM Marshal partnering with various blockchains various AI projects to make sure that like we we come together to give you a experience that that go makes you go wow yeah that sounds incredible uh to have

ai that you can put in so many parameters to find the perfect coin or you know find the project that resonates with you most I'm definitely going to check that out um I I wanted to ask a little bit more about the decentralization part you're talking about the next stages of on Marshall and adding more decentralization now I wanted to differentiate or or figure out

uh with the decentralization in in the data feeds and and the process for example if you had a crypto wallet and in there there's the Bitcoin price on the amount of Bitcoin that you own that price has to come from some location whether it's from an exchange like coinbase uh or maybe it's multiple exchanges so that you have you know in case there's an anomaly on that one

exchange something happens the price is more decentralized but then there's also the process of how that price from coinbase actually gets to the wallet which is the UNM Marshal part itself so is the decentralization both of those parts or how exactly does that work in decentralizing the process more yeah so we already do part of it uh right like when we when somebody purchas

price it's all already normalized across the uh uh across uh exchanges so be considered centralized and decentralized exchanges to give a price feed for users uh but I think like the more decentralization aspect com in where uh the blockchains they have multiple RPGs multiple nodes which is decentralized and the DBS it's it's deployed permission LLY uh across multiple uh

time zones and mult mle instances that's also decentralized I think the only centralized piece that sits in between is the indexer which tracks like which other the tokens you have right like directly you can't access such information on blockchain because everything is stored block by block a simple question like what are the tokens I have uh is is impossible to get

without an indexer so that's where I think such information is is stored in a centralized way so we are opening that up uh for users as a decentralized right so instead of like UNM Marshal storing such data in our stor uh servers we are making sure that even the indexer nodes are decentralized there are like five people who are hosting a same node uh

and then you can fet the data from any one of them and all of them are accurately and they do consensus among each one of them between them to make sure they already have accurate data and and any accurate data any anomalies will will result in their sted amount to be penalized so uh they are incentivized to behave better and they are penalized to penalized if if there is any anomaly in

the data so bringing such a system uh would ensure that there is no centralized piece in this whole narration right like all the way from the node RPC node to what you are seeing in front of your screen uh that's that's what makes uh un martial decentralization unnecessary and the second aspect of it is to remove any centralized Authority in in in uh

building the indexing for the V web 3 right right right now the process is we talk to blockchains the new blockchains that are coming and we work out some sort of uh commercial deal to make sure that like we pay out our server fees we pay our infrastructure Fe fees Etc and and then like we index that chain to make the data available but with de centralization uh the the aspect of any

Central entity would go away so as a new chain let's say you are a new layer three chain built on op stack uh all you need to do is come to un Marshall stack and then just propose hey I want this chain to be indexed and these are the parameters this is the RPC and this is where I want to be indexed and these are the data types I want uh and then anybody can can come and pick up if the

reward is lucrative all right so there is no need for any team to go go and negotiate all economy would be the the entire uh indexing economy would be controlled by March holders like if if the reward March reward is lucrative enough they can index so it becomes infinite scalable solution right so there is no uh Central Services so servers so that we have to worry about

not to uh like not to scale beyond our capabilities uh now uh because it's hosted by individual indexes uh and they can manage a particular service then it becomes super easy to scale uh the entire UNM Marshal network of adding new chains adding new indexes uh without anybody's interven and right so it it becomes infinitely scalable so these are the two major aspects of our

decentralization and we've come up with uh one of the uh I think one of the innovative solutions to build a chain for such a resource intensive task I mean like you can't put like three to four terabytes of data on blockchain and then access uh access that in real time right like you don't have CPU for that but uh thanks to our friends at igen layer we were able to come up with a

solution uh which is ABS based uh blockchain where the um where where the operation or where the execution of indexer happens offline uh but the validation of the data and and validation of the consensus happen on S so this is something we have come up with uh uh in with the with the help of igen layer yeah no I think that's really

important point because blockchains uh each blockchain is a little different in in the cost and the speed but it's very the size of data nowadays even a video file you're looking at like one gigabyte per minute how are you going to actually put that on on a blockchain without it costing a million dollars in transaction fees to transact uh there needs to be some solution like this to

be able to index um and maybe there's a balance of some part is offchain some part is onchain as long as there's a decentralized process and of a fair process of making sure that that information is valid then I feel like that's a perfect solution like on indexing like on Marshall yeah see I think like if you look at blockchains like Solana especially now with the mem coin Mania

uh uh it's just generating uh thousands of transactions every second and like how do you make sure you you keep up to the date uh right so that's where I think such Solutions uh come come to Handy and help us make sure that like we serve accurate data to at all points of time mhm so in this uh decentralization effort in in making the UNM maral protocol more decentralized is that

something that's happening right now there's a process of having community members or anybody that's interested in being a part of the valid validator nodes yes U thanks I think like uh ask the right question at the right time so we are actually running a incentivized test net now we launched probably 5 days ago uh we are dedicating 1 million Marsh uh for which is 1% of our tokens just

for the test incentivizing the test net users um so the the the reason we want is to kind of make sure that like everybody's uh familiar with the uh notes familiar with what kind of notes they are running uh so if you are a marold I mean like the only requirement for you to run any node is uh owning Marsh tokens and like staking Marsh tokens uh to secure your nodes so we

have like three different kind of nodes one is called operator nodes which is where you run the indexer node and the valid data nodes is just securing uh these uh these operators who are running uh Val who are running indexes you make sure that they uh they are behaving properly and another one is called proof reader notes which is additional layer of security uh it's it's like if you

have run any ABS uh on Anon layer you would be very familiar with such uh Concepts so uh it's nothing too different so it's it's same with additional layer of proof reader so uh we would love to uh love I think your community people are are are anybody who's listening to this stock come and run a node on a test incentivized test net and play uh along with the uh test

net and then uh we will be launching main net uh um probably once once we've make sure that testnet both the test Nets are successful M yeah that's really good information and for crypto users that they maybe they've tried defi but they haven't gotten into igen layer reaking they're not developers how hard is it to start getting into something

like these validator notes uh it's actually super uh easy so we have partnered with uh applications like node running applications such as crayon such as node Ops where all you have to do is like click some buttons to deploy a node so they take care of all the infra they take care of uh host in uh validator for you so if you just go with your mass tokens uh we built apps

uh to make sure that it it's super simple for you to deploy respective of whether you have run the noes in uh previously or not if you are super technical person and who want to run the node on your own we have Docker images that you can download and run on your own but uh if you don't want to take that hustle you can use one of the partners uh to run uh node for you so

they'll make sure that they run a a validator on behalf of you for a a small fee um so we made sure that like all sorts of people uh have access to such data and I think if if if uh you see the latest Trend right so running noes has been one of the profitable uh uh like good money earning uh opportunities in the web three now so

there are time of maybe if you look at zero gravity if you look at other apps uh all of them are providing notes uh so it's it's uh super uh uh cool opportunity for people to run nodes nowadays MH yeah I think it's worth looking into um because it gets you more involved in in web 3 and it's different than just holding tokens and and you've mentioned the marsh tokens a few times

uh one to to become validator have some of the Marge tokens but you also mentioned that there was other ways that as the ecosystem is growing and and and being used then the marsh token holders are rewarded could you just touch a little bit more on how exactly the marsh token fits into UNM Marshall and and how it creates a sustainable ecosystem yeah so so far I think Marsh

is a utility token for an Marshal ecosystem uh so far what we've done is uh use like un Marshal token for people who are using un Marshal apis right so they pay using M token to uh get these Services we've built a sustainable uh ecosystem I think that's the way we were able to S sustain like for last 3 years without like leaving lot of tokens

outside uh because we were generating revenue and lot of mar tokens that was uh like that were transaction like uh Supply came back to the March treasury because people were paying in March and we've also burned some of them we have burn close to more than like 700k 800k like close to 800k tokens uh from out of the supply um and right now I think we

are just like removing ourselves from this whole management of uh handling revenue and then burning some of them and giving some of them in taking rewards so because it's truly decentralized the consumers would pay in March and rewards would be paid in March to the node Runners and the the old Network un Marshal chain would be secured using un Marshal token like Mar

token so it plays important role in uh entire decentralization aspect so uh the this it will be secured using Marsh where people need to St at least 25k tokens uh uh and we are looking anywhere between 25 to 30 million tokens to be sted for this purpose and uh out of 100 million and uh and the rewards would be in a similar ratio and all the users

would be uh I think we've built a we build a demand side which is at least like 2 to three million tokens uh com to the uh Network aspect as Revenue uh per quarter so which is like per year I think it's like close to 10 to 15 million right so which is like 15% of the tokens that come back to the uh network uh that will be distributed to the node Runners so uh it's a self-

sustaining model um so we have made sure I think the delay the reason for the delay of network is also this to make sure that we don't uh dilute too many tokens as rewards without having proper demand uh so now I think we feel we we're in a space where all of this can sustain on their own mhm yeah I appreciate you explaining that manahar and if you want to get involved in the

in the validator uh experience and what is the best way to get started in doing that um I think like once you get access to unmarshal if you already have tokens you can just uh come to a community and there tons of documentation video help for you to get started you can just directly go to one of our partner c. Network so to get started uh but if you

don't have Marsh I think you can uh get Marsh either using one of the decentralized exchanges like Unis swap or pancak swap or centralized exchanges like ukin gate uh and mexe and then access um get the data and transfer those tokens to a non-custodial wallet uh so that like you can stake from there uh both holding on E ethereum chain or our BC chain works for us uh and like we

l drop the test net tokens so I think like we have another uh snapshot being taken on 25th of this month uh so if you can like uh participate before 25th you can participate in this uh version of test net if not there'll be another one like this is a two months experiment and after that there be another uh test net before we go to main net that sounds

great I can leave a link to uh the the main platform and the information about the validators uh and any other socials or community links that are necessary as well in the show notes below I really appreciate you taking the time to to teach us and me about blockchain indexing AI uh decentralization more importantly as as well um and I'm wishing you guys all the best on uh

bringing Full decentralization to the protocol and bringing more efficient and effective uh blockchain data indexing to web 3 so thank you so much for taking the time Manara yeah thanks thanks a lot I think this is one of the great conversations I had I think like the questions asked were like super relevant and uh uh hopefully uhan to best of my ability for people to

like understand and then like learn something out of this session

More interviews

Browse all 1,087 interviews

Get new interviews firstThe BlockWest newsletter: markets, AI and policy, twice a week. Free.

Subscribe free