Future Proofing the Financial Rails of the Global Economy I Rodrigo Coelho

We're going to jump right into our next talk. And for this one, I'd like to bring on somebody that's been part of our ecosystem for an amazing amount of time. So, please join me in welcoming Rodrigo from Edge and Node. [applause] All right. Thank you.

Thank you everybody for attending today. I'm happy to be here. First crack of the event. Uh so we're going to be talking today about futureproofing the infrastructure and financial rails of the global economy. Again, I'm Rodrigo Qualos, CEO of Edge and Node.

For those of you that aren't familiar, Edge and Node is the original team behind the graph protocol. Those of you that aren't familiar with Graph, we are an indexing and querying protocol with blockchain data. We've been around since 2018 and we launched a decentralized network in 2021. And since then, we've served trillions of queries. uh the open- source tools being we maintained they downloaded millions of times by developers worldwide and we developed subg graphs which became sort of the industry standard in terms of uh blockchain data indexing and uh as you can imagine over the years we have seen a lot of things over the last few years.

So we're going to talk about some of the lessons from the trenches of data infrastructure. And as you can imagine, we've been through so many it's so overs and we're so backs in these years. We've heard a lot of things. We've been through a lot of problems. We've heard a lot of complaints.

And so we're going to talk about some of those as and discover uh the lessons we've gotten from those as to why we're building what we're building today. So the problems we discovered but first I wanted to take a little sidebar and mention that um if you have you ever thought about the fact that every interaction we have with applications so you're scrolling on X or Instagram it's really a skin on top of a database at the end of the day every everything we do data lives somewhere in in a database and um you know if you think about like Excel it's a database skin on a database at the end of the day so uh everything we do comes back down to data and you know we deal with blockchains we're all talking about data in day so blockchains themselves they were designed for rights um and I give this analogy often sorry if you heard it before But imagine you're writing into a book um and each page is sort of the transaction ledger and each page is blocked in each different page and that closes the block. So you're writing this book but imagine you can't read it. The book is closed. So when a blockchain ships it doesn't come preloaded with software to query it read it.

Uh if you were to buy an offtheshelf database like Postgress it comes out of the box with the ability to read it. Well, that's where uh the graph comes in. We developed the standard to read from databases. So, blockchains weren't really designed for reads. That's where we stepped in and developed a decentralized way to do.

So, we are the read. It's kind of what you can think of when we equate the graph what we do at NGO. So the gap between um blockchains when they emerged, you had Bitcoin, you had Ethereum, the the struggle to get re readable data off them and build applications on top of these blockchain which at the end of the day are databases slowed industry adoption first and then our original founders saw this problem and developed subgraphs as a means to solve this problem. And so subgras are the thing that bridged that gap and it unlocked the defi area sub the defi era. So how many of you here were around in 2021?

Okay so 2021 defi summer unis swap all those things. This may not be known to many people but the graph was behind D5 summer. So the graph was powering all those platforms and dashboards. uh we provided that data and made that whole thing work and behind the scenes you know we were keeping the infrastructure running and the the scale and the pace at which it happened surprised us all was sort of unexpected but that was kind of the beginning of like product market fit for uh blockchains and what they could do and what impact they could have on the world so we're happy to be part of that as you can imagine we learned a lot from that And you know, we've been building ever since. We haven't gone anywhere.

Many people have said, "What are you guys doing? What are you gonna ship something?" So, we uh we're happy to still be around and still building solutions for that. So, as I mentioned, you know, unis swap uh odd these tools were were backed by subgraphs and it was perfect for that time of period for uh defi's experiment experimentation phase. So it provided the data that was needed for these systems to operate.

But then the industry caught fire and started to scale um and scaling was set to fundamentally challenge infrastructure. So think about it. We've got these new high throughput chains, right? That wasn't expected before. Um Salana came around and now now things are moving towards faster and faster chains, more data.

um transaction volumes dilute um we've you know with like uh be coins and NFTTS and all these things the scale at which things accelerate and uh we fast push these systems to their limits um and data volumes grew exponentially. So you know when you have blocks happening every 200 milliseconds with a lot of data the data adds up and we get into like pabyte scale you know histories of of blockchains right and so that becomes challenging and then new requirements emerged you know over these years like the industry was maturing what's the big trend now enterprise right we're hearing enterprise enterprise enterprise so enterprise scale business needs emerged you know with the change in legislation, the opening of um financial institutions wanting to transition antiquated systems into uh blockchain system and new financial rails. So they have new needs, desires, wants. Um and so we've got new realities to deal with. So enterprise businesses have totally different requirements that were unexpected here to here to before.

Uh so we were dealing with sort of the blockchain developer kind of audience uh in the prior years you know developing DeFi apps decentralized apps and so now we're dealing with a new reality the largest financial institutions in the world wanting to bring systems online and it's a totally different audience. So what what do these enterprises need right they need security auditability performance and flexibility these are non-negotiables and so we spent years of learning about how to bring on data to world systems. So what do they actually mean? Um so you know the world's largest institutions they need data they can trust query and govern just like other systems. um onrem private cloud manage.

So as we're speaking to these large financial institutions on premise is something that has emerged as a major requirement right many of them are running private chains they have private financial transactions they don't want exposed so the ability to run these systems within their uh secure environments within their data centers is an important um factor that we've learned. So that becomes a new requirement that we hadn't thought about before, right? [clears throat] Data sources must be compliant, auditable, auditable, and trustworthy. So when you're dealing with um in the case of DTCC, like a quadrillion dollar of transaction settlement a year, they are under major scrutiny by regulators and there are major financial pen penalties uh if things are wrong. You cannot screw this up.

So um compliance is a major requirement of these financial institutions and with that things need to be auditable and trustworthy. Verifiability is a huge factor and uh they need to ensure that the data that's sitting in their environment matches what's on the blockchain. So there needs to be proofs for that. And then true multi-chain capability. So new chains are coming online all the time.

It seems like another week, another chain, right? And so there needs to be a quick and easy onboarding for these new chains as they come online, like not a whole new implementation for a complex onboarding for each new chain. So that's a requirement too. So meeting these needs required us to rethink from ground zero and develop something entirely new knowing these new uh requirements. So we built AMP.

AMP is the first blockchain native database. So what that means is we built this you know with our years of experience from the ground up from the first principles knowing these new requirements and saying okay starting today um we're in a new era we have new demands new needs new uh use cases new users what would we build today knowing what we know now so that is AMP so with AMP you could build publish remix smart contract data sets. So we call this composability. So think of it like Lego blocks. You can mix and match data from different places, create your own views and materialize them and uh publish them to a decentralized registry.

Just going to get through some quick benchmarks. I know this is no short, but I do want to mention some of these are pretty spectacular. So over two times faster than other full-blown databases. So it's storage efficient almost six times faster than BigQuery on public data sets 100 times faster um than competitors giving sub subsecond data freshness. Uh so if this were the case uh that would be like a 1 second load with AMP while you'd be waiting 100 seconds or almost two minutes for your data to load with other competitors.

um and over 4,000 times faster back fills compared to traditional RPC. So [snorts] these are just like spectacular numbers and 4 million events processed per second which like I mentioned earlier when you're dealing with a quadrillion dollar settlement you need to have super fastput data processing. So AMP meets all these needs of new reality. So now we're going to switch gears a little bit and talk about AI and agent adoption. Um it's growing exponentially.

We saw this coming. Uh we had early discussions um earlier in the year about where things were heading and given our experience of data knowing that AI agents were going to need data we looked at how we could participate uh in this new reality. So current AI infrastructure it can't keep up with verifiability visibility and payments and it doesn't support the agent economy scaling on change. So let's look at some of the problems here with the agent economy. Zero observability.

It take agent actions, right? Like what is my agent doing? Um it's uh running on its own. There's no unified way to manage payment permissions. Uh no data provenence for what agents act on.

And so to accelerate the growth of agentic economies, there's been several new stands. How many of you are following kind of what's happening in the AI agent space? Okay. Um so many of you are have already heard of X42. Uh this just blew up to the broader world in the last three or four weeks.

It was launched back in like March. It's since taken fire, but this is really an open standard that defines ways in which agents transact with each other. Um, it's been a huge unlock for the space. Um, we were contributors to that with a a deferred payment scheme that allows for batching of transactions and micro payments that are well below the threshold of the x2 standard. So, we were happy to participate in that.

Google's A2A region standard. This is for agents how a standard for how they communicate with each other. Um so we've been working closely with them and ERC 80004 which is an open standard developed by the Ethereum community for reputation and discovery. Um in a world of you know agents that just have wallet addresses you need a way to determine like the trustability of who's on the other side of this transaction you're about to take. with um X402, you just simply make a payment.

And so you're you need to be sure that the agent on the other side is trustworthy and is going to do what it is setting out to do, what it says it's going to do. So there were two missing trust layers needed for the agentic economy scale. Need data integrity. So, how do you know you can trust what an agent is doing or what information is making decisions with if you don't have verifiability and observability? And economic accountability.

So, how can you place trust in agents if you have no way to track, monitor, and control their financial transactions? So X402, A2A and ERC804 are the public standards. But what's needed is a unifying system here. A unifying agentic standard paved the way for scaling the agentic economy. So if only there was a solution for this, I wonder what it could be.

enter amperand. We um we spent the last uh almost year building this and we are opening it up today. It's at number.ai. So, it's built on X42's foundation.

It gives humans the control panel they need to observe and control Jetic payments. and it connects data integrity with economic accountability. So you may have um A16Z put out their state of crypto report a few weeks ago and in it there was this Gartner estimate that by 2030 there would be $30 trillion in purchases that would be made or influenced by AI agents. Okay, so that's 5 years from now. today is almost zero.

So we're talking an exponential scale in how all of us in this room are going to interact with commerce. So everything that we do on our phones today or on the computer tapping, clicking, paying, all that is going to be handled by an agent, right? So think about you know the greatest minds in the world like Johnny I working with OpenAI about new interfaces for computing and for agents like they're talking about voice interaction so we'll be speaking with an agent that knows all of our preferences that has the ability to transact on our behalf and you know even here I was thinking through what a future would look like. I got to the airport, Uber didn't work. My credit card wasn't working.

I had to someone said download Cavali. I had to go big cash, right? It was a mess. So, imagine get off the plane or even before you get off the plane and say, "Hey, get get me a ride to Montre Hotel and then find me a restaurant, book eight people. You know, my preferences, I love Thai food, right?"

And it's just great. I found a new place. Book everything handled. You're you're talking to it. It's not like you're clicking around in your phone, looking on Google Maps, looking at reviews, right?

Think about all the interactions you're doing on your phone to just handle these things. So, the future is within 5 years, we're talking and this is all happening for us and it's the payments are mediated through explor 2. Um, and so we envision a future where we're going to have swarms of agents. Like each of us will have dozens of agents that handle different aspects of our lives. Working in airlines here, a restaurant, right?

They're going to be kind of task specific rather than generalized. You might have a general agent that's like your main agent that knows you and then that agent will go find other agents to do for you. So, amperand is a wallet for agents and a dashboard for payments. So, real simply, it's a visual um view of what your agent forms are doing, right? You can see how much they're spending, what transactions they're doing.

You need control. If you think of an agent like a teenager with you gave a credit card without a limit, you want to give your teenager some spending controls, right? you can only spend a couple dollars, you know, $10 a day, like $2 per transaction or whatever it is. Um, you don't want them going out and like training all your your accounts. Um, so real simply, it's it's dashboards to be able to view do all the analytics, control everything.

And so like within your agent forms, you can see agent by agent what they're doing. I mean, it's a pretty simple idea, but it's needed in order for this thing to scale. And in a world where we may have a billion agent wallets floating around doing all these different things, we need to have visibility and control and auditability, all the things that I mentioned before, the visiting to scale. So, AMP proves what happens, [snorts] which is uh the first product I mentioned. Aerson proves to be paid.

So to summarize um you know we've been through as I mentioned a lot of we're so we're so back I think we're on the we're so back right now the graph and into node are creating the next era onchain infrastructure these are new realities and I think you know we've been through about a decade of the Ethereum 145 we're now getting to a point where the things we're working It's like the rubber hitting the road. We're we're going to be utilizing these tools that affect us every day. If you think about like financial transactions, so stock and bond trades happening on these these rails. These are real world things, not just limited to a small group of niche people. And then agents interacting.

everyone here is going to be affected by all this um this new uh change that's happening. So we at graphic engineer happen to be part of all this wave that's happening. So here's my contact info. Um if you'd like more information on the airport codes happy to connect with anybody here you guys a second to take those photos and That's all.

Thank you very much.