Intro to Sahara AI: The Blockchain Built for AI

[Music] Hello everyone, welcome back to another stream and uh in this stream we will be talking about Sahara, a new and

interesting blockchain which focuses on AI and AI developers but it's not just limited to that it's open for everyone.

So joining us today from Sahara Labs we have Jesse Gild who leads product at Sahara Labs and he will be taking us all of us through in and out of Sahara

blockchain and how all of us can build and deploy our AI agents or data sets on the blockchain itself. So I will pass it

on to Jesse you so that you can greet our audience give a little bit introduction about yourself and take us all through what Sahara is and how all of us can start building on Sahara.

Thanks Hill. Yeah it's great to be here.

Um really excited to join this session.

Um so I'm Jesse uh at Sahara AI. I lead

the blockchain and protocol side of the product development here. Um bit of background on myself. I've been in the web 3 blockchain space for the better

part of eight uh years of my career. Um

and uh I've been at a few different web three companies including Avalabs and Zero Hash. And I'm really excited now to

Zero Hash. And I'm really excited now to be kind of at that intersection of AI and uh crypto which seems to be um one of the the really I think exciting

spaces to be in right now. Um, and so what we're building at Sahara is decentralized AI uh solutions for open

and programmable ownership of uh what we call AI assets on the blockchain. And

you know at its core um you know we're helping AI developers and uh data set creators to tokenize and monetize the the work that they do uh and the

contributions that they make towards the development of AI in a secure and transparent way on chain. So that's the overall um background. So happy to dive

into Hill unless you have any you know initial questions happy to cover those off. Yeah. So Jesse I read somewhere

off. Yeah. So Jesse I read somewhere that Sahara is a AI native blockchain.

It's a AI native network. So for all of us who are not that much fluent in with the AI space because you know all of us

are just Dian over here. So would you like to explain in simpler terms that what does this term or what does this phrase AI native means in general and

what makes Sahara a AI native network?

Yeah. No, that's a good question. Um so

AI native really means a blockchain built by AI developers for AI developers. Um and so it's it's really

developers. Um and so it's it's really um you know a chain that uh really understands you know the process of AI development and the needs of AI

developers um and building that really you know into the core foundations of the blockchain and the smart contract protocols and solving for the needs of

of that audience. Um so Sahara is really a purpose-built blockchain. It's not

meant to be a general purpose blockchain. Um it's you know purposely

blockchain. Um it's you know purposely built for the tokenization of things like data sets uh agents and models and

enabling more uh what I call a monetization and providence layer that allows you to um basically track ownership of these different types of

assets on the blockchain and enabling uh fair ownership and you know equitable ownership of you know people's uh contributions to the development of AI.

Um, so this isn't like a blockchain that's really simply um integrating with AI agents or being retrofitted with, you know, a agent uh, you know, uh, DeFi.

It's really infrastructure that's being built, um, you know, to help open up this what we call sort of AI economy um, and create a more um, transparent and

more equitable environment for people that want to build AI. So just a little bit of background on Sahara. So Sahara

has you know at its core AI in its DNA.

So uh unlike some projects and other you know web 3 companies out there um Sahara is really starting off as a web 2 AI company and now moving into the world of

web 3 and blockchain and bringing you know that that that deep understanding of AI with it. Um, and Sahara is backed by, you know, a very strong, you know,

visionary investors across both the AI as well as crypto ecosystems. And so this lets us kind of merge, you know, the best of both of these worlds to tackle these problems that we're focused

on, which are things like, you know, the opacity of um AI development and, you know, being able to enable better attribution uh to address kind of what

we kind of consider to be the the extractive economy of today's AI ecosystem. And so just kind of I guess

ecosystem. And so just kind of I guess breaking down like what what we call like decentralized AI is that's sort of a you know a fairly dense concept. So

you know we can kind of unpack that a little bit but um what that means is you know data sets models agents becoming traceable ownable and programmable on

the blockchain. And so what you get with

the blockchain. And so what you get with uh decentralized AI is an understanding of who uh contributed to the development of AI models uh proof of you know who

actually contributed to the creation of data sets. um you have um the ability to

data sets. um you have um the ability to incentivize as well as enable um onchain payments and um using our smart contract

uh framework creating composable uh AI assets things like onchain licenses and and you know usage uh policy enforcement on chain um and then finally being able

to you know transparently have the rewards and the the royalties for the success of these different assets on chain um basically flowing and routing to the

creators um in a transparent way programmically on the blockchain.

Monetization and participation is you know one of the sort of I'd say like key uh factors of of what we're trying to enable here. And so right now like one

enable here. And so right now like one of the core issues is that AI development contributions um often go unrewarded. So you know you have these

unrewarded. So you know you have these sort of um unsung heroes of AI development. The people that are out

development. The people that are out there basically um curating data sets, labeling them, annotating them. Um those

are people that put a lot of time and energy into you know improving uh these you know foundational models as well as enabling the fine-tuning of models that

often don't get their their fair share of the benefit of AI. And so what we're trying to enable here is allowing data set curators um to earn fees through our

programmable licenses and then um have the the developers of different models and agents also um be able to earn more directly from their products um on a per

inference basis using our onchain payment systems. What this allows is more fair and transparent development and uh contributors are are basically

able to you know better um realize you know the the upside and the benefit of their uh their efforts when it comes to you know things like data set creation as well as agent development. So

basically, yeah, basically what this means is, you know, if somebody's building on top of something that you've you've created, um, you know, you will, uh, not only be recognized and not only

will you have your efforts attributed to, you know, these different, um, agents and models that are including your data, um, but you'll also be

rewarded. Got it. So let's say that I I

rewarded. Got it. So let's say that I I have a AI model or let's say that as simple as a data set not even a model

and I want to tokenize it be it a model be it a data set so how would I go on to tokenize it and uh start earning rewards

from that uh will I have to go to a specific platform will I have to deploy some information in a smart contract or how does that look like? Say you are, you know, uploading a climate data set

for um, you know, the benefit of someone that wants to create a, uh, you know, a a climate science agent. Um, basically

what you would do is you would have your data set. You can go to the Sahara

data set. You can go to the Sahara developer platform, which is one of our main applications, and you can register that on chain. And by doing so,

basically what you get is an an NFT that represents your ownership. It's like a um you know title or a deed that basically indicates that that you're the

creator of this particular asset on the blockchain. Um and then what that allows

blockchain. Um and then what that allows you to do is later create a license NFT which allows you to basically set the policy for usage. So we have different

types of licenses um for commercial non-commercial use um and it allows you to basically define the usage rights uh for that particular data set. Um and so

as users are consuming your data set and embedding it into their projects whether it be you know an agent or a rag pipeline um people will be licensing

your data set and that will be triggering uh a distribution of payment.

So you can set the price for your product on our marketplace and then people can go and and browse that marketplace and purchase um the various data sets. So the creator of that

data sets. So the creator of that climate data set have their royalties basically streamed back to them. Um and

we use different types of uh you know uh token um token structures to enable this. So we use obviously ERC 721 uh

this. So we use obviously ERC 721 uh NFTTS to uh show the root ownership of these different assets on chain but then

we also use um timebound and usagebound access tokens. So ERC uh 3525 um tokens

access tokens. So ERC uh 3525 um tokens for being able to basically set those specific um policy access features. And

then we also have what we call contribution NFTts. So um if you're

contribution NFTts. So um if you're somebody that's participating in the creation of a data set using um our data

services platform, you can um basically earn uh what we call a tokenbound um or soulbound NFT um which is our contribution NFT. And once you've

contribution NFT. And once you've contributed a certain number of data points to a data set, um that allows you to basically stake your contribution NFT

and to staking pools and then you would be then able to earn um a portion of the royalties or the revenues that are associated with that particular data

set. Um and so each asset that gets

set. Um and so each asset that gets created has its own um wallet using an ERC uh 6551 um tokenbound account

wallet. And um that's basically how

wallet. And um that's basically how people would be receiving their rewards.

Um and so you can think of you know this whole uh layer as sort of this um you know monetization layer um that's uh

basically like programmable equity in you know the success of these different data sets and and agents. And so we have different um methods for tokenizing both

data sets and agents. But the front door to our protocols um are basically our developer platform where you can go and register these assets um as well as get

access to our SDKs as well as our uh agent builder and then also our marketplace which is really um you know you can think of that as like hugging face on chain where you're able to

basically browse and select different types of data sets um and models to to build your agents with. And Jesse just one follow-up question on that. You

mentioned that you folks have different types of licenses as well. So is there a open license like some somewhat similar to MIT license where yes if someone

wants to open source their data set they can just open source it and anyone can use it without paying any royalty or anything. Absolutely. Yeah. Yeah. So we

anything. Absolutely. Yeah. Yeah. So we

we are not just focused on monetization solely. We also want to help you know

solely. We also want to help you know you know proliferate the uh you know open- source and the MIT licensed uh you

know assets as well. So absolutely yes we we do have that as as a particular option when it comes to minting licenses. Um we have multiple different

licenses. Um we have multiple different open source license types that you can select from. Um but then also in

select from. Um but then also in addition to that you have the ability to set a commercial license or a remix license. um that gives you the option to

license. um that gives you the option to basically create a monetization policy for your asset. So definitely both are

are important parts of the uh important parts of the the uh the whole flow.

Got it. Got it. Yeah. I mean that's very good to have because uh you know we as developers loves open source. Yes. No,

totally. I think like you know a big part of what we're trying to do is obviously we want to allow people to contribute to the open source community but then also for those people that you

know do this as a career and make money um which there's there's a lot of people out there that are contributing to data sets um as you know their full-time job.

We want to make it easier for them to also be rewarded for that and have their efforts be um you know not not go unnoticed and also um not be extracted

by you know other um actors out there that that aren't necessarily um you know um trying to just exploit Yeah. Yeah.

basically aren't um you know recognizing that as well as you know having the proportionate amount of you know shares and rewards flow back to them. The

similar thing there is one question from KHL. Uh since this is very related to

KHL. Uh since this is very related to the current topic that is why I want to bring it up right now. So he asked that data sets can be big they can be huge

right. So are these data sets stored

right. So are these data sets stored like onchain or is there any other mechanism to do that? Yeah. No they're

not stored on chain the data sets them themselves are are stored uh offchain using what we call our our vault system.

So we have our own um decentralized storage solutions that that's where the the data itself will be hosted. Um so

yeah that's a really great question. Um

it wouldn't make uh it wouldn't be feasible for us to be able to store all of this data on the blockchain. The

blockchain itself is really more of the um you know orchestration and reward layer for people um to have the provenence of their contributions be

tracked. Um but yeah the actual storage

tracked. Um but yeah the actual storage of the data itself is is offchain. Just

a quick look at you know some of the products that we're building. So as

mentioned you know we're really focused on the uh the development of AI and building solutions for AI developers on chain. So really starting with you know

chain. So really starting with you know our data services platform which is sort of at the root of you know AI value without you know clean uh labeled

trustworthy data um it's it's really difficult to get great models. So this

is like the sort of raw material and you know the the you know what I sort of call like the oil um of the sort of AI

industrial economy. Um and so our data

industrial economy. Um and so our data services platform um is already live and it's a basically a decentralized peerreview platform where contributors

can um basically uh you know provide data points um to different data challenges. So there's different data

challenges. So there's different data tasks that get requested by our enterprise partners that are looking to um build out data sets and our platform is built as sort of a decentralized

peerreview system where contributors engage in a peerreview process and they ensure the data integrity and reliability for through um sort of a collective validation process and we

have incentive mechanisms built directly into the platform um where you know there's basically a structured reward system that keeps the contributors

rewarded and motivated to maintain a high degree of you know data quality and so that aligns the efforts of the labelers and the annotators with the people that are requesting the data.

Right now um we have over 100,000 contributors on the uh Sahara data services platform. um we have about a

services platform. um we have about a 92% uh data point approval rating which means that overall the the data set quality is very high but then there is a certain number of data points that get

rejected um and yeah this is the the sort of I guess the the foundation for the Sahara ecosystem which is creating

this um platform that turns raw data into tokenized uh assets and um this is what is sort of forming the informing the training of you know the next

generation of AI models and agents and so this is the first product that we're launching we're also launching a AI studio so just before before you go into

AI studio I have a question related to data services platform so you mentioned that there is a peer review which

happens for data and that is for high accuracy of data so how do you ensure that these reviewers themselves provide

accurate and reliable reviews and reliable data on reviews. Yeah. So we

have different ways of doing that and um you know one of them is through what we call like honeyotss. So we have we inject certain uh questions and certain

tasks that we know are essentially bogus. And if people are basically

bogus. And if people are basically providing us with the wrong answers to these questions, um we can essentially detect whether or not their

contributions are honest and of high quality. and that will basically lead um

quality. and that will basically lead um their contributions to be you know sort of tombstoned and not included in the overall data set. So we have different

um sort of I guess safeguards and guard rails that we've built in um on top of different roles. So you have roles of

different roles. So you have roles of reviewers and validators and it's just their job basically to review these data points and so um through you know having these different types of rules as well

as um you know checks and balances in the form of honeypotss um we are able to more or less ensure that there's a higher uh standard and a higher quality

of data um and you know basically if you're participating in the AI studio you're doing that in order to get rewarded and you know if if you know the

system detects that someone's trying to game or farm the system by not contributing high quality data, it will wean that out and you just won't earn

rewards. So, um yeah, basically our our

rewards. So, um yeah, basically our our our tokconomics are there to basically keep those incentives in mind for the people that are participating in the the

uh the data services platform. Um and

then we're also you know right now working on introducing some AI based um sort of adversarial auditing. So ways of basically programmatically checking the

quality of the data and having AI um be able to challenge the quality of the data. So that's another way that we're

data. So that's another way that we're currently um looking to sort of just ensure that that that data quality remains high. So on top of that we are

remains high. So on top of that we are also building an AI studio which is sort of like you know GitHub um for AI development um and it's uh you know a

studio that allows you to perform all these onchain actions um it allows you to uh register your data sets um license your models license your data sets as

well as agents and um you know basically um do all of the things that AI developers need to do to to build their products. Um a couple of like the core

products. Um a couple of like the core pillars of the AI studio are the data hub. So this allows you to um discover

hub. So this allows you to um discover and access and and monetize high quality data sets as well as our model hub. So this allows you to basically access different uh you

know open models like you know llama um deepseek um as well as you know other fine-tuned models that might be derived from um you know the the data sets that

we have on the marketplace um as well as custom agents as well. And then our compute hub which we we partner with um you know the best uh you know compute

providers out there like leptin uh hyperbolic um and so basically for the you know you know once you have developed your uh AI agent and you want

to you know have u you know run that you can utilize our compute hub and basically get access to inference directly from the platform as well as

pay for that inference.

um using the Sahara token. So, it's

basically, you know, a one-stop solution for um building as well as launching AI agents um you know, all in one place on the blockchain with built-in providence

tracking using our smart contract protocols as well as you know, incentives through our token. uh in

terms of you know royalty sharing and revenue flows. This is a little bit of a

revenue flows. This is a little bit of a busy chart but you know the the crux of this really is our programmable uh royalty system ensures that that value

is is always flowing upstream to the creators. Um and so if somebody has um

creators. Um and so if somebody has um taken your data set and embedded that into a rag pipeline or has used it to create an agent and that agent is having

inference uh you know uh API queries run against it that those a portion of those royalties will flow back to the creators of the data sets. And so royalties are

being enforced on chain. And we use something called our our royalty graph to track um you know all of the sort of

um you know regressions up to the the upstream owners of the the different assets that are being contributed uh for these different agents and ensure that

ownership is always being recognized and royalties are flowing um you know to the to the owners. Right now we are in test net. So we are uh we've just launched

net. So we are uh we've just launched what we call Siwa. Siwa is our sandbox environment for people to start going and actually testing out our uh you know

our our protocols as well as testing out our applications. Um so this just went

our applications. Um so this just went live um a week ago and um it's basically starting off with data set tokenization.

So um we've had you know a tremendous amount of activity on SEAW already.

We've had um I think about close to about a million um you know uh different uh touch points on chain. So people

actually registering um the um you know their data sets on chain and minting ownership NFTts. So it's been uh a

ownership NFTts. So it's been uh a tremendous amount of uptake um on SEAW already. And so that's where we're

already. And so that's where we're starting right now is um the data set tokenization enabling data sets to be tokenized. People can basically go and

tokenized. People can basically go and mint an ownership NFT and um they they register that using what we call our global registry. Um and so this is you

global registry. Um and so this is you know starting to ensure providence and attribution and they access this through the Sahara dev portal. Um the next um

roll out and the next release uh for Siwa is going to be our licensing and royalty monetization protocols. So the

next iteration of this is going to be being able to mint a licensed NFT and automate the the distribution of fees um as well as be able to start to have um

you know royalties start to stream back to their royalty vaults. Um and so this is you know the first step in sort of empowering creators to monetize their AI

assets. Um next uh you know iteration of

assets. Um next uh you know iteration of of SEAW will be agent tokenization and um contribution NFTts. So we're going to

be enabling people to tokenize agents um through um our agent builder. And so

what that basically allows you to do is set up multi-step workflows and multi-step pipelines which can include multiple data sets as well as prompts um

using our agent builder and basically track that full lineage of um assets. So

you can think of this as like your Lego builder um for an agent and for each data set that gets um added to that agent you'll have contribution tracking.

So the people that have created these data sets um will have their contributions tracked using contribution NFTts and um that's the next step of allowing data set creators and data

providers to have their efforts rewarded um through the creation of of um agents.

And then finally later in um July what we're going to be doing is fully open sourcing all of our protocols. right

now. Um they're they're basically built into our front end, but we're going to have a fully permissionless uh open- source protocols that will allow anybody to basically build on top of these

protocols um with, you know, different uh smart contract templates. And um this will really be an invitation for all web 3 builders to basically come and start

to um you know build on top of these protocols um and start to build them into their own set of DAPs as they um as they want to. Um and so that's the

current CY test net roadmap. Um and then on top of all these things we are going to be rolling out lots of improvements to our applications. So, we're going to

be um iterating on our developer portal as well as um releasing our AI marketplace which as I mentioned is kind of like you know uh an onchain version

of of hugging face. Got it. So Jesse on the road map probably a dumb question but uh I just have one question that since the data tokenization or data set

tokenization is live now and the agents tokenization will be coming up in like you know coming months as you mentioned.

So during the time of uh agent tokenization when that is live will we be able to use the current data sets which are already tokenized or during that time some like the network will

upgrade and we will have to post new data sets. Um no yeah that's a good

data sets. Um no yeah that's a good question. So the data sets that get are

question. So the data sets that get are get that are getting um uploaded now will will be usable um after we push these new uh changes in terms of you

know agent tokenization. So you know all the data sets that are getting uploaded today um they will be usable moving forward. Um and and same with you know

forward. Um and and same with you know once we actually move to mainet we will be um you know making sure that there's a onetoone match of the uh NFTts that

are getting minted on Siwa and ensure that that providence and that ownership also gets tracked on mainet. So you

won't be losing um your ownership rights even once we you know migrate over to to mainet with Sahara. So that's a really good question. Um, yeah, so you can

good question. Um, yeah, so you can think of this as right now like we're just rolling out the ability for people to tokenize data sets and start to earn rewards. Um, but then uh with the launch

rewards. Um, but then uh with the launch of our agent builder, we'll be then opening up our protocols to also tokenizing more complex uh agent workflows as well. Basically, I guess

just to sum it up in the time that we have, you know, you know, building with Sahara um is really about uh creating this transparency and earning potential

for people that want to build AI. Um you

know, the Sahara protocols are um you know, built for transparent attribution.

Um as well as allowing people to just have better control and ownership um over their data sets, over their agents.

Um and then creating this you know monetization layer um for people that um are building on chain. Um we are also looking to push the boundaries of you

know DeFi as well. So while Sahara AI is an AI native um blockchain built for AI developers, we're also um building some really really cool stuff in the DeFi

space um that's um you know in keeping with the sort of AI native uh sort of mission going to be driven primarily by onchain agents. So we're really excited

onchain agents. So we're really excited for that as well. And um yeah, this is just a quick look at you know the the information for anybody that wants to

actually start building their DAPs on Sahara. So this is you know the the

Sahara. So this is you know the the network name Sahara test net. You can

see the RPC UR URL here um and our chain ID. So developers are are more than

ID. So developers are are more than welcome to start you know coming and deploying their their DAPs. Um, and we are, you know, really looking to, you know, work with different, you know,

projects that may be building onchain agents. Um, and so we're definitely, you

agents. Um, and so we're definitely, you know, open to, um, you know, exploring these types of projects with developers out there. Um, as well as, um, you know,

out there. Um, as well as, um, you know, we're starting to actually incubate a few projects as well. So, we're really excited about that. We have an incubator program. So if there's any projects out

program. So if there's any projects out there that are interested um you know we have different um programs out there to help kind of facilitate uh the growth of this ecosystem and and help builders uh

get set up on Sahara. So thank you so much. I mean this was it was really

much. I mean this was it was really great to be able to share more um about Sahara with your audience and uh we're we're thrilled about the partnership with Quicknode. Um you know you guys are

with Quicknode. Um you know you guys are you know best in breed when it comes to you know web3 infrastructure. So we know that um our infra is in amazing hands

and uh we're excited to to grow the ecosystem with you guys for sure. Thank

you so much Jesse for doing this and explaining all the audience members about Sahara uh how they can build their own you know data sets and deploy their

own data sets and uh AI models on chain and uh you know start earning from it.

For those of you who are watching this as a YouTube video, feel free to drop those questions in the comment section and we will make sure to pass them to Jesse and the Sahara team. So in the

meantime, if uh like in the meantime while you're thinking on those questions uh Jesse, if you don't mind, I myself

have uh two general questions around Sahara.

So the first one is could you briefly uh explain or talk a little bit about

how easy it is for non web3 developer who is you know uh AI AI native developer to

get started with Sahara get started on Sahara SA for example developers who are very familiar with uh open AI APIs as

entropic APIs and uh even vice versa for web3 developers who are not very familiar with AI stuff. So

like vice versa question for non web3b3 developers to get started on Sahara and for uh web3 developers to get started on the Sahara

how easy is it for them because this is kind of a hybrid situation where it's a very good collaboration or meeting point

of AI and blockchain tech. Yeah. No,

that's a great point and it's something that you know as as the product team um we we're really focused on this because um as you rightly mentioned right like

we're kind of at this intersection of of web 2 AI and web 3 and so we need to make um the experience um appealing and and approachable for for both audiences.

So um you know one of the big things for us is you know creating um an a user experience and an environment that's familiar to AI developers. So um one of

the things that's important there is abstracting some of the complexity of web 3 away from the experience. So um

you know using things like account abstraction um and gas abstraction to sort of smooth that out. So um we're working with partners to uh deploy um

you know 4337 account abstraction as well as you know our own relayers to facilitate you know gas sponsorship. So

when people want to interact with the blockchain using our interface it's it's not you know something that they have there's a lot of learning curve there.

Um, of course for people that are familiar with web 3, they won't have any issue, you know, using their MetaMask wallet and they won't be in any way thrown off by, you know, different prompts for, uh, paying for gas or

confirming um, transactions. But for

somebody that's not familiar with that, it could be a bit of a, you know, um, a jarring experience and create, you know, a bit of friction there. So what we're really trying to focus on there is just

you know smoothing out the the onboarding experience as well as just overall as you're actually interacting with our protocols of you know setting up your data set and tokenizing it um

minting an ownership NFT or a license NFT each of those different uh points along the journey will will you know be smoothed out using um account abstraction and gas abstraction. So

that's one of the points there. But then

you know when it comes to uh web two developers that are familiar with um you know primarily maybe working with a APIs um we we have our own APIs and you know

these would be following things like the open a open AI API standard. Um so

getting started with actually building applications is is really really easy for AI developers. um you know, if they've built AI applications using

OpenAI's um APIs, um all they need to do is just switch out that endpoint address and their app should be up and running on Sahara. So, it's a um it's a really,

on Sahara. So, it's a um it's a really, you know, smooth transition from using your web 2 uh developer tools and then sort of making that transition over to

Sahara. So that's the way we're really

Sahara. So that's the way we're really focused is just like on that that onboarding and overall like the tooling itself um in the the AI developer studio

will be very familiar. Um it will look and feel a lot like what you see when you're using anthropic or open AAI. Um

so yeah that's that's a big focus for us on the product side. Um and then you know for web 3 developers as well um I think it it presents a really really

clean and approachable um you know interface that again if you're building onchain agents all of these different tools should be very familiar for you.

Um, and then, you know, if you're wanting to get deeper into the weeds on our smart contract documentation and start actually, you know, deploying your own contracts, um, you can just, you

know, basically use our documentation and, um, it will have all the, um, the familiar sort of quick start guides for for web 3 developers that want to go down kind of their own path. And we'll

also be, you know, releasing our own SDKs as well as a Cly tool so that people that want to build um, you know, in in a more web 3 native way um,

they'll be able to do that. On the

closing note, I just have one closing question. So uh would you like to share

question. So uh would you like to share or if you can remember any interesting project or use case which you have seen

on the Sahara SA testnet or any interest for a particular use case if you would like to share something. Yeah

absolutely. So we have some really cool projects happening. Um one of them in

projects happening. Um one of them in partnership with myshell. Um so we we're doing some really cool uh data challenges with myshell. Um Michelle is

helping us sponsor some um you know data tasks. So you know people that are using

tasks. So you know people that are using our data services platform um can go on there and find different data tasks that will you know allow them to earn um earn

Sahara points which will then qualify them for airdrops later down the road.

Um so that's a cool initiative that we have in partnership with my shell. Um, a

couple other really interesting ones that we have is we're building our own uh DID de FAI uh agentic uh trading uh

solution. So, we're creating our own

solution. So, we're creating our own trading co-pilot that's going to be integrated with an MCP server that allows people to basically use

autonomous trading bots um on different uh sex and and DEX platforms um starting with uh Binance. So, this is a really

cool one. It's using Sahara um a agent

cool one. It's using Sahara um a agent uh builder to to create the agent and then it's also being trained on data sets that have been um basically

tokenized on the Sahara uh platform. And

then lastly we are um working with a project called Axgrad which is a model evaluation and benchmarking um uh platform. So you can think of this as

platform. So you can think of this as sort of like chatbot arena on chain. Um

so people that are deploying their own um chat uh deploying their own agents on our blockchain um can basically have those uh you know be listed on this uh

chatbot leaderboard and it's basically an open model evaluation where people can basically evaluate the efficacy of these agents and models and there's also going to be a tieback to uh our

protocols as well as incentive uh systems for that as well. So people that are building really effective uh agents will um you know also be um have having

those efforts be rewarded through our our token layer. So those are just a few I think of the really cool examples of of projects that are building on Sahara right now. But um yeah, we're we're

right now. But um yeah, we're we're really excited for the future in terms of, you know, what kinds of other really cool uh agent uh use cases we can facilitate and we're really excited to

see just like the the crazy amount of uptake that we've seen on Siwa. So the

the number of data sets has been um really encouraging. So, we're excited

really encouraging. So, we're excited for once we launch our marketplace to see um you know, a really strong supply of of great assets um there waiting for

everybody once it opens up. For everyone

who's watching, be on the lookout for those new upcoming Sahara tutorials which will be in collaboration with Sahara and Cutnote team. So thank you so

much everyone who joined us and uh a big thank you to you Jesse from myself and the entire quick node team for joining us and uh you know enlighting everyone

about Sahara. Yeah thanks Ail it's been

about Sahara. Yeah thanks Ail it's been a pleasure uh having the session with you and um we're really excited about the partnership with QuickNote. So thank

you so much for the opportunity to to share.

Thank you so much and uh thank you so much everyone. We'll see you in the next

much everyone. We'll see you in the next one.