Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
Building Trustless and Unique Agents On BNB Chain
Recorded: Feb. 25, 2026 Duration: 0:50:51Space Recording
Full Transcription
Thank you. Thank you. Thank you. Thank you. Thank you. Hey everyone, let's wait for a few minutes for everyone to get in. Thank you. Okay, good morning, afternoon, evening, depending on where you're calling from.
Welcome to BNB Chain's space again.
And once again, we are going to talk about the highly anticipated vertical AI.
I'm Walter from Business Development at bme chain we are one of the largest blockchain out there with um you know very diverse ecosystems including artificial intelligence so today we have
you know very interesting uh founders and developers here with us to discuss about
building trustless and unique agents on the any chain so trustless right you
know the word trustless arm I think it arises from you know nothing arises more about trustless
than reputation being on chain right transparent and immutable and that is from the you know
recent protocol ERC is usual for so to everyone, BME Chain has this protocol on our chain as well.
And for unique agents,
agents can be tokenized.
And this way you capture the intelligence on chain.
And that can be done
through our very own application protocol, BAP578.
So do check those out.
And BME Chain chain we have many
different layered narratives on AI you know last year we saw a lot of payments
growth I say layered narrative because those narratives don't die away they
they constantly layer and build up the AI space on BMB chain you know we
recently had the open claw hackathon as well a lot of people join in and right
now I think in the past few days we have seen thousands
and thousands of projects registering on bmb chains 8.004 and also minting their bap 578
uh tokenized agents so that is fantastic you know i look forward to growing more and before i start
i'll highlight that you know nothing said here is financial advice as always and you know everyone
should do their own research before interacting with any web3 protocols so yeah let's start by introducing our guests so maybe starting with
chat and build crystal please hi walter thank you so much and it's always a pleasure to be here on
bmb chains space so my name is crystal i'm the founder of chat and build and we've also invented
My name is Crystal. I'm the founder of Chat&Build, and we've also invented the BAP578 protocol,
which basically allows you to have a verifiable resume for agents, which tracks identity and
receipts for what they did, under what permissions, and what outcomes. You can actually go to nfascan.net
to actually see all of this happening in real time. Every transaction is also cross-referenced
on BSC Trace. So at Chat and
Build, our mission is very simple and anyone should be able to build anything instantly just
by describing it. We work very closely with some of the top AI companies in the world, particularly
with OpenAI and Google as well. We received grants from them as well as Entropic. We've been featured
on Entropic. So we shipped our coding agent last August and now we're building the next step, ChatChat.
So ChatChat is actually a platform where anyone can build a personalized agent in real life with persistent context and memory without needing to understand frameworks or code.
And a lot of the BAP578 technology is going to be incorporated into that.
So just a quick kind of scan on literally NFAScan.net as of now,
we're already tracking 24,983 verified agents out of 27,588,
which is massive and all of this had just happened in less than a month.
And it's really processed over 150k transactions,
thanks to the amazing B&B chain community.
So I'd like to also echo what Walter has said,
nothing here said is financial advice
and please do your own research. Yep, that's amazing growth. Thank you for sharing. Later on,
we will touch a bit more on, you know, actually, Chat&Build co-authored BAP 578 with BMP Chain.
But before that, let's introduce our next guest, Alex from DGrid, please.
Alex from DGrid, please.
Hi everybody.
This is Alex from DGrid.
And thanks for having me here.
And thanks for the support from BNB Chat.
And we are building,
like there are two main products from DGrid AI.
One is AI Gateway.
So we are providing a single API
and by using this API,
users can access more than 200 AI models with one API.
And this is the infrastructure.
And based on this infrastructure,
we have our Dory is a AI agent,
and it will analyze your request
and call the different APIs.
As you know, like different large language models,
they have something strong area,
like they are good at doing different stuff.
So the agent will analyze your request and call the
strongest large language models to fulfill your needs and also we are building a lot of fun stuff
on a bnb chain for example bap 578 and like you will see all of the products in the near future
This is not a financial wise. So
please
Keep in touch stay tuned and try our products. Thank you so much
That is great. Thank you for the introduction next up read from Sinex
Sinex can you hear me?
Is it just me or I can't hear Sinex?
Yeah, I cannot either.
Let's wait for like 10 seconds.
Hi everyone.
This is Tony Sai from CNX.
It's my turn, right?
Yeah, we can hear you now.
Awesome.
Okay, thank you.
Okay, so firstly, thanks to BNB Chen to have this chance to have a discussion with experts from different areas.
So this is Fonisai from CRNX. We provide several auto agents that can support also the individual developers and also the
business companies to build their own agent or this kind of super agent applications based
on our platform. Actually currently we already have several different
fields of all-time MBM.
So we already launched our platform since the early of 2024
and currently we already have like millions of developers
from different countries, like more than hundreds
of different countries are developers and also business companies based on our marketing platforms.
So I'm very glad to have the chance to discuss with you guys about these trustless and the unique agents on BNB chain.
And since we launched our applications on BNB chain
around December, 2025,
and currently our applications already reached
the number one new launch project on VMware Chain.
So also welcome you guys to share our applications and also our blockchain platforms.
Thanks to you guys.
Thank you so much for the introduction.
We will now move into the questions.
So let's talk about the core principles of unique and trustless agents. So how do you define trustless agents and what makes it
fundamentally meaningful to have agents deployed on a blockchain? So perhaps you can invite Crystal
from Chat and Build to have a discussion on this. Yeah, thank you so much, Walter. So I think
a lot of people are now talking about trustless agents, right?
I think for me,
it doesn't mean that the agent
is actually magically good,
you know, because, you know,
as we have seen, right,
the agent space is evolving so quickly
and there's so many new things
happening every single day,
you know, once with OpenClaw
and new open source frameworks.
And everyone here is trying to
build out their own agentic projects.
But I think where trustless agents actually is important, it's actually about having a system
that is verifiable, even when you don't trust the operator. And that's where the magic of the
blockchain comes about. So for me, at least, and from the perspective of BAP578, a trustless agent
actually has three different properties. The first thing is that it needs to have identity that can persist. So it's the same agent across time and across
applications. Number two, a trustless agent needs to have permissions that you can inspect. So what
it can do and what it can never do. I'm sure some of you have already seen that, you know, some
agents have had access to wallets, you know, they can drain wallets and misspend and wipe out your
emails up. These permissions are actually really important as more and more agents come into play
for us, right? And I think third, but last but not least, I would say is receipts and provenance,
because you need to know what the agent has actually done, under what constraints and what
actually changed. So on-chain actually really matters here because it gives us that shared
ground truth across parties. And in Web2, you see that trust is actually the brand, but in an agent
economy, agents are able to interact with strangers, cross organization boundaries, and
transact frequently. So we need that neutral substrate where identity authorization and
action history can be independently verified.
And that's where BAP578 came about.
And we co-authored this together with the BNB chain tech team late August last year.
And now it's kind of like, I think, swept into a new life of its own,
thanks to a lot of the open-claw stuff that has happened.
And we've seen a lot of these signals already happening
through NFA Scan, the various projects building on BAP578. So I would say in summary, trustless agents is really about verifiable
learning and receipts that are becoming normal behavior.
Yeah, thank you so much. Very well said. I think, you know, permissions are very important
where, you know, agents are still at a very early stage. We need to have gut reels to
make them really trustless. Yeah, thank you so much. And next up, Diggit, please share your views on this.
Diggit Kuo- All right, okay.
So in my point of view, trustless agent, AI agent that operate without relying on any
centralized party.
So this is very important. And this is also what blockchain brings us,
the decentralized stuff.
And the trust-less agent,
they act based on code and on-chain rules.
They are now controlled by a team, company or server.
So I think this is the basic definition of trustless agent.
And the real meaning of putting agents on blockchain, I think it is very important because before the AI generation here, like all the player
are individual users.
But right now, like there are many AI agents on blockchain.
So the real meaning of putting agents on blockchain is the first agent's logic cannot be secretly changed or shut down by any team or any centralized companies.
And the second one is transparency.
So everyone can verify what the agent is designed to like how how the work is doing right now and the last
but not least is uh native ownership so uh the agents can directly uh own assets or in inter
interact with d5 project or nft project so uh like just like I said about it's another
player in the web3 world.
So yeah, that's it.
Thanks.
Yeah, thank you so much
for sharing. Definitely agree
on all these. And yeah, I think
producing humans is difficult, right? But producing agents
is so easy, yet
each agent can be so unique and contribute
in many different ways so
i think the growth will be very exciting and uh quite exponential looking forward to that
uh next up uh cnx maybe can share your views on this okay okay thanks uh bm chen so this is my
opinion so uh i think a trust in this agent is a system that
should be operates without requiring users to trust in any
third party. Okay, so the central authorities or individual
entity. So it should be a mechanism and principles are
based on several different key ideas. The first one is the decentralization.
So the agent runs on,
should be run on a distributed network
instead of a single server.
No single party controls the entire process.
So I think that the BNB chain
provides us a very good environment.
The second one is that the cryptographic proofs
and rules. All actions should be governed by predefined code and cryptography and the
conscious rules for the human promises or authority. The third one is the verify ability. Every operation is transparent and veritable.
That's very important. And no anyone on the network and the users do not need to start
and they can check and verify the results themselves. And the fourth one is that the no intermediary,
that means it removes the middle term, middleman and transaction decisions or service
are executed automatically and directly between parties once conditions are met.
So this is my opinion.
Thanks again, Daniel Lee Chen.
Yeah, thank you so much for sharing.
So yeah, I think the transparency part of the chain
allows agents to be clear.
We can be very clear of what are the agents on chain
and that creates accountability.
I think ultimately all of the guests
is highlighting this area where how blockchain can actually
improve the AI side of things because a lot of LMs could be black boxes and so on.
So having accountability of what the agents actually are is very important.
So yeah, this is the main topic of today.
And I think it will be very important as the cornerstone for AI growth.
Next up, I did mention Chenabute, Crystal's team.
Crystal herself has co-authored the BAP 578.
So maybe we'll take this opportunity with the developers and community here to maybe let Chat and Build share.
What are the key messages you would like to share with developers and the community on utilizing BAP 578?
Yeah, thank you so much, Walter. So BAP 578 is, I would say, a labour of love from a lot of us because we really looked
at, I think at that point in time, when we started working on it last year, we didn't
think about it being a standard that would just be static and was designed for what the
models were like with solid 3.5.
We did it when Anthropic was 3.5 and GPT was 3.5.
We built this so that it will evolve with the large language models
in the next six months, eight months, ten months.
I think that was one very important thing.
So I think with BAP578, the first important point I want to say
is that everything is open source and it's available on our GitHub.
We have live PRs as well.
So you're welcome to take a look at everything that we have there and also contribute to it and make this really a community-driven open standard.
I think that's the first thing.
So my message here is that BAP578 is not treating it like an NFT of a new label.
It is an Asian standard.
So I have four points actually on
this part that hopefully will be good for developers to know. So the first point is that
we want to make the receipts boring and correct, right? So as your agent evolves, the learning and
the state snapshot has to be deterministic and verifiable, and that's the trust promise. So
we've had a lot of really great BAP578 projects already built
and one of them is BORT as well.
So that's also a great project to check out
and see what they have done there with learning agents.
The second point is that we have to treat upgradable logic
as your biggest attack surface, right?
So if agents can actually bind logic contracts,
you want very default safe patterns like loudest,
emergency resets,
and audit trail events. And we've already had community PRs hardening this for us from a
security standpoint. And we're really, really thankful for that because we want to work on
this together, the community. And that's what we should converge on as a baseline.
So the third part as well, I think that I to highlight is that Verified needs a real bar.
So verification should actually mean that compatibility tests and reference checks are there.
Otherwise, it just becomes marketing.
Last but not least, which is a very important point,
BAP 578 is composable and interoperable with other registries like ERC 8004,
transport like MCP and A2A, payments like X402.
So products can be opinionated,
but that doesn't mean that standards are also in a silo.
Standards should stay composable and we all have to work together as one big ecosystem.
I mean, I think there are a few of these people that go,
oh, you know, X402 is base, you know, ERC-8004 is Ethereum.
No, I mean, the whole blockchain space right now,
we all have to work together
because it is a tiny fraction
of what the AI market is, right?
So we are in very, very early days.
And if all of us can work together
as a holistic standard,
then, you know,
we can bring the rest of AI into blockchain.
And that's what we want to do.
Yes, thank you so much.
Yeah, very upset.
Like just the hardware maker for AI and NVIDIA, its market cap
is larger than the entire blockchain space. So definitely, we are here to build our area of the
AI overlapping with Web3. And I think a key point, takeaway is definitely the protocols are
complementing each other. So BAP578, you tokenize the agent. It is data on-chain.
And then with that locked in place,
you can register using ERC-8004.
So the agent is registered,
and then the community, the ecosystem,
provides the on-chain repetition points for it.
So this whole thing actually flows
to create a proper sort of agent hierarchy
and discoverability for users out there.
So that's great.
I think a great point so far.
We are moving to starting to talk about actual use cases now.
So maybe back to CianX.
Perhaps you can share some of the stuff the agent can do
is autonomous trading.
So what are your views on fully autonomous AI trading?
And do you think this is the future
or it has to be developed very carefully?
OK, yeah, thanks. And do you think this is the future or it has to be developed very carefully?
Okay, yeah, thanks. Okay, you know, currently the OpenClaw is very popular and I also spent several days in the Chinese New Spring Festival.
I found that there's still a lot of challenges that make the agents to run automatically
smoothly.
So my opinion is that how to ensure the agents or the large geolongering models to run smoothly
and reliable. My opinion is that currently the LLMs alone cannot guarantee
sufficiently stable, smooth, and consistent cushion or autonomous agents. I think that
to achieve the momentum of the trading or other scenarios is that we need a structured mechanism.
The first one is that the modular architecture separation,
that means how to speed the system into several different layers.
That the LM is only only one key condition,
and also that the determining execution layer
also very important to help us to fix the logic,
like also the verification and the tax code computation.
So how to ensure that the deterministic ground health and also the confidence is very important.
I tried by myself how to use the strict scheme mask and the ortho-functioning calls and the state machine to limit the LLM output, math, and behavior. And the feedback loop and the
self-correction also very important. So how we correctly implement a real-time validation and
error detection to make sure the LLM to produce the inconsistent or invalid output,
make sure the agent automatically corrects itself without human innovations.
So I think still that we have a longer road to make sure everything runs automatically, smoothly, and makes the consistent output is still a very big challenge.
That's my opinion. Thanks. Yeah, thank you so much. Consistency and quality of outputs is very
important. From the chain side, we do feel also that maybe security is also a key thing where
there's always a balance between, you know, people get excited for autonomous tasks,
but there is always some give and take with security,
but we are slowly typing the gap.
And I think these are great pointers for developers to take note off.
Next up for DGREED, we have a question where we want to know, you know,
how do you think developers or BNB chain can exist in building up and progressing large language models?
For example, the classic ones like OpenAI,
Cloud, DeepSync and so on.
Do you think blockchain can help with the,
open sourcing the data maybe,
or leveraging blockchain any other way?
Feel free to share your points.
Yeah, of course.
So I believe blockchain and BNB chain can fix two big problems for LLMs.
One is like it can bring real data to users. And the second one is they will like the blockchain technology also BNB chain can give fair rewards to all of the users,
includes like agent users, agent providers, LIM providers, or like data providers.
So blockchain can prove where data comes from so for open large language models worthless or repeated data is a
very big problem but on-chain on-chain checks makes the make the data trusted and clear at least
useful for you know building powerful LLMs.
And the second is decentralized rewards.
People who give compute data or model improvements
can get paid fairly.
As I said before, like model providers or data providers,
like model providers or data providers,
like they contribute a project or a large demo model,
and they should receive rewards.
So blockchain, also BNB chain can solve this problem.
Also like BNB chain can be the payment layer for ai that's why we are building
on bnb chain and data model usage and ai services can be priced on blockchain and settled safely
so that's my opinion like that's why like it is very helpful for solving all of these problems.
Yeah, thank you, Alex. I would like to highlight the point on, you know, I mentioned that on BNB Chain, AI has layered narrative.
So one of it is actually AI data ecosystem that Alex covered where, you know, blockchain, you can make micropayments across the world. You know, when you send a payment from a wallet to wallet, it's not dual bounded.
So this way users can actually provide data for AI agents or models to train on and then
decentralized incentives in the form of tokens can be distributed. This way you get uncensored
information to train AI agent models. So yeah, I think great points.
We will move into, we talked about the logic
and then what's going on now.
Let's talk about the future.
What are the future real use cases?
AI is growing very fast though.
So let's discuss on this.
What are the real world use cases that you believe
that's gonna happen in the future
or observing is going very fast now that will demonstrate maybe the clearest product market
fit for agentic AI on-chain on BNB chain.
Maybe Crystal from Chatt & Build can share first.
Thank you so much, Walter.
This is something, again, I'm very passionate about because, you know, as I've said earlier
in our chat, right, AI is evolving so fast. So it's
so important that you don't just have static agents, you also have learning agents. And at
BAP578, we piloted this idea that you could have Miracle Learning Agents, right? So what's really
interesting today is like, you know, if you go into NFAScan.net, you're able to see that, you know,
people have already minted 27,000 vertical learning agents.
And this means that people aren't just minting, they're also evolving these agents with verifiable
learning snapshots. So I think just to answer your question a bit more directly, I think PMF,
product market fit that is, shows up where agents are able to cross boundaries and trust is
expensive. So there are three areas I think think, with the use case-wise.
So the first use case I see is agent commerce and microservices.
So basically, agents paying for data, compute, and execution
with budgets and receipts.
Very, very solid use case, I would say.
Something that all of us can use.
The second thing is, you know, ops agents for teams, right?
So in the future, you're not just going to have
an army of agents working for you. Your agents are going to manage their own sub-agents as well, right? So in the future, you're not just going to have an army of agents working for you,
your agents are going to manage their own sub-agents as well, right? So the agent is going
to propose, the humans are going to be approving, and there's verifiable history. And that is the
infrastructure that we are building for. Third part is a bit close to my heart because I do believe
that, yes, you know, payments are important, infrastructure is important, but in reality, all of us love real social and gaming experiential experiences, right? So consumer games
and social, I think, are where agents are able to evolve, and this is where the non-fungible aspect
becomes very intuitive. Every agent can be a non-fungible agent, and that's why we call, you
know, BAP578 NFA, non-fungible agent,
because every agent that you make is very unique
and that's why it's non-fungible.
So that's why it becomes intuitive
because it is a combination of identity,
progression, and provenance.
So I'm very excited about the space,
and I think these are the three very key areas
where you can have PMF for on-chain and agentic AI on BNB.
I mean,
BNB chain
has so many
great categories
across gaming
and social.
you know,
all of this
can be applied
to agents as well.
Yeah,
100%.
I think at some point
we are going to see
our own AI agents
having AMAs themselves
and then,
you know,
developer agents will be listening in to learn how to improve and then you know developer agents yeah we'll be
listening in to learn how to improve and then they repeat this loop over again and again and
I don't know what will happen but definitely progress. So next time we have an AMA Walter
I'll say is it Walter or is it Walter's agent? Yeah yeah I mean that might happen yeah so yeah
thank you so much for sharing. Next up D Degree, maybe you can share, you know, your views on the future product market
field.
Well, like, I personally really like auto invest, auto investing and auto trading.
Like, in the past, you know, like I have to go through all of the different projects,
DeFi projects to find the highest interest.
But right now, like, you know,
I don't have to go through all of the aggregator, DeFi aggregator.
I just like let AI do all of, AI agent do all of the work.
So he will find the highest income,
highest profits aggregator for me.
And he will also like automatically
doing the staking behavior for me.
That's one point.
And another one is auto trading.
Like for example, on BNB chain,
like there are a lot of good projects
and there are a lot of, you know, decentralized exchange
with good trading volume.
So I don't have to, you know,
create a centralized exchange account.
Like it helps me like skip all of the process
and the AI agents just, you know, auto trading on the decks.
Like by right now I'm losing money.
Well, it really depends on the basic layer,
which is large language model.
Like what model are you using?
And what's your strategy of trading?
But like when you find a way to trade
and AI agents will help you earn a lot of money from trading.
And this is one point I like. Another point I like very much is you can trade.
Sorry, I think there's a misclick.
Oh, can you hear me right now?
Yeah, I can hear you.
Oh, yeah.
Yeah, the last but not least is the AI agent can wait for me when I'm sleeping.
So when something big happens, he will make the decision for me so I can stop losing my money that's why I'm so excited about auto invest and auto
trading that's it thanks yeah thank you so much for sharing I think you know AI
has a lot of application and what you mentioned is actually a superb
aggregator and at real time as well.
And not just aggregating, like you mentioned, the trading strategies.
Like it can even read your blockchain history to understand what kind of investor are you.
And then it can determine the strategies automatically.
So, you know, maybe you are very conservative.
So a 10% drop, it would have stop loss, something like that.
So looking forward for ai to
reach that level and then i think you'll benefit uh many people uh down the line so yeah um we
will move to uh cianex uh could you share you know your views on future product market feeds
okay yeah okay thanks uh work uh although uh cian Although Cialex is the number one AI agent based
crypto training agent on Aster, we still face a lot of challenges.
Although the LMS have become powerful enough, but the high quality data remains a major challenge for
building stable, smooth and trustworthy agents. My opening is that how to structure high quality
data makes agents more stable and reliable. There are several different dimensions or different ways that can maybe in the future
to make sure the agents' performance are better. The first one is that the clean, structured,
and verified data, how to remove noise and duplicate the basis. And the second one is that domain-specific and task-aligned data is also very important.
So how to train and fine-tune LLNs with data tailored to real-world agent scenarios.
And the third one is that continuous data validation and feedback also
are very important. Because you know how to establish a real-time validation mechanism to
check data quality during the agent execution, how to use the users and system feedback to
system feedback to internally, intuitively refine data set.
You know, I have more than 20 years experience in the AI experience with the traditional
business customers like the retail manufacturers. retail, manufacturers, I found that the data quality still the biggest challenges for them.
So my opinion is that how to find the fact that the process to make sure the data sources,
the data quality to make sure the LM also the agents performance better is still the key.
Maybe we still need several years to make this happen. Thanks, Wouter again.
Yeah, thanks. Thank you so much for sharing. Definitely. I think there are many other different
like future product market feeds for on-chain on BNB.
I look forward to many, many different innovations coming on board.
So we have talked about the maker side of things now.
Maybe we can talk about your own projects.
You know, what is the future milestones that you guys are preparing for?
And so that our audience can keep an eye on your product on BNB chain.
So maybe chat and build, you can share first.
Okay, thank you, Walter. So we are building a lot of really exciting things at Chat and Build,
and we've already actually announced this. It's called Chat Chat. So basically, Chat Chat is your
very own personalized AI agent platform, and we're co-developing it together very closely with the
OpenAI team and the Google team as well. So we're
super excited about that. We've opened up the waitlist already and we'll slowly start getting
people in to try the platform very soon. So we're just hardening up some stuff right now. So keep a
look out for that. But yeah, we want to make agent creation feel very normal and safe for everyday
users. So I think the second thing that, you know, we are doing a lot of stuff as well with BAP578,
and we want BAP578 to mature into a real community-run standard
with that shared safety defaults and verifiable receipts.
And we've really started to see lots of activity around that
across the B&B trade ecosystem.
So we're happy to be able to do that, you know, with the community.
So the one-liner here, again, just to kind of recap,
is that BAP578 is the verifiable resume for agents, identity plus receipts for actions and
permissions and outcomes. And you're able to track all of this, not just on our GitHub at
chat and build, but also on NFAScan.net, which was actually developed by the community, by someone here, Ben Toda, who's listening on this space as well.
So that can give you a snapshot of all of the verified agents and projects
that's happening on the BAP 578 standard.
So we welcome all of you to contribute to it.
So the next step for this is to make verified meaningful
and to make the tooling more obvious.
So things like compliance kits, explorers, and consistent security patterns.
The third thing of what we need from the ecosystem
is we would love to see more builders shipping real use cases
and contributing back via PRs and reviews
because we believe that standards get stronger
when people try to break them in public and fix them together.
So that's our roadmap.
We're shipping the creation layer, we're hardening the trust layer,
and we're scaling adoption really safely.
Thank you so much.
I like it when we're building innovations,
but the word safety and harden has been mentioned.
I think that is really a very important emphasis.
As we innovate,
safety is always the most important
for any industry like a blockchain industry. So yeah, that is great. Thank you for sharing industry like, you know, blockchain industry.
So, yeah, that is great. Thank you for sharing.
Next, DGREED, you can share more about the future of your product.
All right, sure.
We are preparing something fun for AI agents on BNB chain.
We will launch open trustless agent ecosystem using BAP578.
We will have a marketplace for all of the AI agents building with BAP578 on Degree.
So the marketplace, any developer can build and launch AI agent.
And like if some users are using the AI agent, the agent provider will receive rewards.
Also like the users will pay for it or they will like have a chance to use it for free.
So for users, get ready for a smarter, easier Web3 experience.
And for developers, get ready to build on DGrade
and with BAP578.
That's it.
Thanks.
Thank you so much.
This is amazing.
Like what we have spoken about, the community gets together
to build, Ben, Alex, everyone, you know,
are putting in developments around BAP 578.
I just want to highlight that, you know,
tokenizing agents, right, you know,
for the audience to understand,
it actually brings demand and supply to the agent.
You know, some concept could be AI agents
having a token taker,
but the performance of the agent doesn't matter.
But if you tokenize it, it's like an NFT thing of it
as each AI is unique
and it is not performing well.
It does not command high demand.
And with a high demand, there isn't a good price, for example.
So it is based on merits.
So this is what we want to achieve, right,
in Web3 and free world.
So thank you so much.
And next up, CLX, please share about the future of your product.
And next up, SianX, please share about, you know,
the future of your product.
Okay. Thanks again, Wot.
Okay. Again, this is Fanchisei from CLX.
We provide the multi-agent architecture and platform for also the individual developers and also the business companies.
That's why we provide a solution, the market agent platform is that while we face the
complex scenarios, the single agent is not sufficient enough to tackle the complex real-world tasks.
So our solution is that we try to build a multi-agent platform to resolve the complicated
real-world issues or problems.
So the first one is that we will solve the complex problems and break it down into clear, manageable
subtexts, then to specialized agents. And also, each agent focuses on specific capabilities,
such as planning, reasoning, execution, verification, or complications.
So maybe this is the best solution to resolve the real world challenges.
And also, you know, the agents should interact and negotiate and complement each other through a structured communication protocols,
creating synergies that no single agent can achieve.
And welcome you guys to experience our
the CO Next multi-agent platform
and also build your own applications
based on our platform.
Thanks, you guys.
Thank you so much.
Yeah, I think, you know, the strongest framework
is where agents work with each other, like you mentioned,
and then having the framework and platform for them to do that.
It's great, especially on a blockchain, right?
When payments, et cetera, it just makes sense
using the chronological and immutable transparency nature of blockchain.
So thank you so much, three of the very strong AI projects on BME Chain sharing about all
these different important pointers on BME Chain.
Once again, I will re-emphasize this chat is great.
A lot of things are being learnt here, but nothing said here is financial advice and
users should always do their own research before interacting with Web3 protocols.
Do follow BNB Chain, of course.
We will have many exciting events around AI.
We've just done hackathons, and then there will be another hackathon in India, and so on.
And of course, the developers matters.
Do follow Chat&Build, DGrid, CNX, and all the projects being developed on BME Chain,
so that we can support and grow the AI space together.
So yeah, do support each other and continue to build and build.
Thank you so much, everyone, for the time.
Cheers, everyone.
Thank you, everyone.
Thanks, guys.
Bye-bye.
Thank you for having me here.
Bye-bye. Bye. Thank you.
