Kristin Myers: Good afternoon from the floor of the NYSE. This is Startup to Stock Exchange. I’m Kristin Myers. Now, every AI company you’ve heard of is built on top of something a lot less glamorous: the plumbing that moves data from where it sits to where a model can use it. And that plumbing is a business, a really big one, with big incumbents. And my guest’s company spent years building it for blockchains, a small, specific world. And now they’re taking what they learned and pointing it at everybody else. But first, let’s take a look at what they’re building. The road from an idea to the opening bell starts right now.
AI runs on data, but as companies generate more of it than ever, that data is often scattered across different systems, formats, and networks. For developers, finding it is one problem. Making it actually useful is another. That’s the problem Rodrigo Coelho and Edge & Node want to solve. Edge & Node helped build The Graph, infrastructure designed to make blockchain data easier to access.
And now it’s taking that experience beyond Web3. Its new platform, AMP, is designed to help developers access and work with data from different sources without having to build all of that infrastructure themselves. And the timing could be critical. The AI boom is creating an explosion in demand for fast, reliable data, and as AI agents become more powerful, they’ll need access to more information from more places in real time.
But here’s the challenge: Edge & Node is entering a data infrastructure market already filled with some very big players. So can AMP turn its Web3 expertise into a mainstream data business and scale enough to eventually make the leap to the public markets? From Startup to Stock Exchange, Edge & Node CEO Rodrigo Coelho is next.
And joining us now here at the desk is Rodrigo Coelho, CEO of Edge & Node. Rodrigo, thank you so much for joining us today. So I want to start with The Graph. Talk us through what that is. What problem did it solve, and why was blockchain data so hard to get out in the first place?
Rodrigo Coelho: Yeah. So it was around the 2017, 2018 era that there was a big boom in ICOs, and Ethereum was coming up. And there was a problem, in that blockchain data is not really stored in a table or spreadsheet format. And so what was happening was developers were having to reinvent the wheel every time to extract data from blockchains and make it available for applications they wanted to build on top of the blockchain.
And so the founders of The Graph, I was the first employee brought in, in 2018, but the original founders were developers themselves, and they thought, this is crazy. We’re having to stand up a cloud database to make a copy of the blockchain, and then you’re doing that over and over for every application that needed to be built. And they thought, okay, let’s just make a standardized protocol. We will bring in all the blockchain data, and then we’ll make specific databases for application developers, called subgraphs.
So we kind of invented the space and saw that it was a big problem that everyone was having. And then from 2018 to 2021, and even till today, it’s become a very important part of Web3 and blockchain infrastructure. And The Graph has served trillions of queries over the years and still serves billions of queries today. There’s a decentralized network of computers around the world that are serving this data to all kinds of applications. And so it was really solving this big problem, which is that blockchain data itself is not really stored in a filtered, organized way. And this is bringing order to that chaos, and then making it fast and available for high-demand applications that need it.
Kristin Myers: So let’s talk about your role, because you took over as CEO last year, I believe. Did you walk in and decide that anything needed to change? How would you define your tenure as CEO?
Rodrigo Coelho: Yes. Well, as I just mentioned, I was the first hire, so I was there for the last eight years. And so I was kind of the fifth Beatle and helped the founders, you know, with scaling the company. I was more on the operational side. And so when I stepped in, I had all the context of the entire history. I knew everybody on the team.
And so the big changes that I made were saying no to things and cutting back. And basically, we had to focus. The company at that time was doing a lot of things. We were focused on many different areas. And so this was about retrenching and turning to a singular focus, which became AMP. And that’s our enterprise product now that we’re selling to tier one financial institutions, still working with blockchain data, but really turning our focus in on that.
Kristin Myers: So let’s talk AMP. What does a developer do with it that they weren’t able to do before?
Rodrigo Coelho: Yeah. So we’re still dealing with extraction and organization of blockchain data. The big difference was what we did before was specific to a certain use case. And what we found, as we started working in 2021, this was open-source software, and we started to notice that financial institutions were actually downloading it and playing with it, and, you know, major financial institutions. And so we started to see, okay, banks, stablecoin issuers, tier one financial institutions are now leveraging blockchain data.
And we learned that they had different needs than the small kind of Web3 blockchain developer at that time. And so they had existing data pipelines. They had data stacks that needed to be plugged into, and their needs were more on analytics, on audit and compliance. And so we spent a couple of years in R&D, and then we refocused. The main thing was leveraging SQL, which is Structured Query Language. It’s the most widely used language among developers, and it allows you to write simple statements to get data from, you know, tables of data.
And so it’s based on that. It’s based on open standards. It’s based on an open file format, so it allows for high compression, so it saves a lot of money in terms of infrastructure. So it’s utilizing best-in-class technologies and state of the art, and it’s really emphasizing this pluggability and adaptability to be used for many different use cases. So rather than being sort of application-specific, this is broad and can really be flexible enough to fit into many different implementations and use cases.
Kristin Myers: So speaking about that, you’re used to a world, and you’ve sort of built in a world, where a lot of the data is public. But a lot of corporate data is private and not publicly accessible. So what did you learn, really, when dealing with those types of transfers?
Rodrigo Coelho: Yeah, yeah. So interestingly, Ethereum and other blockchains, as you know, are kind of open public data. And as financial institutions started to come online, they had a demand for private data. And so there are blockchains out there that are specifically addressing that. The kind of leader in that space is Canton. It was formed by a consortium of banks, and what it allows for is the ability to separate data that’s private to you, and it allows two institutions to transact privately without everybody seeing what’s going on.
So the beauty of what we’ve built is that it’s compatible with that. Given that our software can be run on premise or within certain closed cloud environments, they can install and utilize our extraction tool and platform. And because they have the keys to unlock that data, they’ll be able to just see their own private data without exposing it to everyone. So having the ability to run it themselves gives them access to just the data that they can see.
Kristin Myers: So let’s talk about, I guess we’ll call them, some of your competitors, right? You’re not the only company out there in this space that does this. There’s also Snowflake, which I think a lot of folks have heard of. There’s Databricks, Confluent, cloud providers. They’ve also been doing a lot of this work for years, and they’re huge. So talk to us about what Edge & Node provides that perhaps they don’t, that might be of interest to a potential client.
Rodrigo Coelho: Yeah. So the companies you mentioned are really focused on off-chain data, and they’re dealing with data lakes and data warehouses. I would view us as complementary to them, because we’re kind of the world’s foremost experts in on-chain data, given that we invented the space. What we build actually plugs right into Snowflake and Databricks. So many of these institutions that we work with already have these data warehouses that they use, you know, with the companies you mentioned. But what they don’t have is the ability to pipe that data in, and that’s where we come in.
So it’s almost like an extraction or a filtering mechanism to get it into the standardized format that they’re already used to utilizing in these big data lakes. And so their data engineers can still do their same analytics in their Snowflake or Databricks environment. The data coming in gets normalized and adjusted to match what they’re already used to.
And so the benefit is that they don’t have to write this from scratch and build it themselves. And they can own their data pipeline, where their on-chain data just comes in with off-chain data. So say it’s market data combined with on-chain data. And you have a lot of financial institutions that are, you know, doing structured assets or tokenized assets that need off-chain data, for example, and they want to combine that. So this just allows them to buy rather than build. They can plug right in. It’s sort of like a Lego brick, right?
Kristin Myers: Plug and play.
Rodrigo Coelho: Plug and play. Right.
Kristin Myers: So I want to ask you about the customer base. You’re mentioning a lot of financial institutions that are using this now. I’m curious to know what kind of expansion you’ve seen, even within the world of finance, especially as a lot of banks, for example, are now moving to doing things on-chain.
Rodrigo Coelho: Yeah, this is a global trend. So we’re seeing this. I spent the last, you know, my tenure as CEO traveling the globe, going to different financial conferences in every region. And it’s really a megatrend globally. And so what I view is that the entire plumbing of the financial world is being rebuilt.
So for example, just a quick example: when you hit send on a wire, or you send a Venmo or a Zelle, we don’t think about what happens behind the scenes, you know, for that money to get to that other individual or institution. And so what’s happening is everything behind the scenes is being rebuilt, or there are parallel kinds of rails that are running on stablecoins. And so stablecoins allow for, you know, lower transaction costs, instant settlement, fungibility, without needing to run over decades-old infrastructure.
Kristin Myers: I mean, even ACH, I’ve heard, still uses like a daily FTP file transfer in a batch.
Rodrigo Coelho: It does.
Kristin Myers: And our Startup to Stock Exchange viewers have heard about this over the last couple of weeks, as we’ve talked to a lot of companies attempting to solve that problem of connecting new plumbing with very old plumbing.
So why is getting at data so hard? Well, picture a midsize company. Its customer records sit in one system, and then its payments sit in another, and its website behavior sits somewhere else. Its support tickets are in a fourth place. Now, each one of those was bought at a different time by a different team, and each one stores things in its own way.
Now, let’s say somebody wants to build something that uses all four. Well, traditionally, that means pipeline software that’s going to pull from each source, reshape it, and keep it current. So engineers build it, engineers maintain it, and it breaks whenever any of the four systems change. Now, surveys routinely find data engineers spend more time moving and cleaning data than analyzing it. And that work is the market.
We’re back now with Rodrigo Coelho, CEO of Edge & Node. So, Rodrigo, let’s talk about, you know, the AI shift. What do those AI systems need from data infrastructure to essentially do their jobs?
Rodrigo Coelho: Yeah. The first thing I’d say is verifiability, and that’s one thing that not all current data pipelines offer, which we do. It’s kind of the garbage-in, garbage-out problem, right? And so we’re using the blockchain as a source of truth, and then you’re bringing a copy of that data, you know, into your environment. And so how do you ensure and trust that that data matches, right? It could have been modified. It could have been tampered with.
And so what we have is cryptographic verifiability against the chain. And the beauty of blockchains is they use hashes and mathematical proofs that then tie each block together. And so we have the ability, to an auditor or regulator, you can actually have a proof that everything is 100% accurate and complete. So that’s one aspect: completeness, accuracy, verifiability, which ensures the trust that institutions need when they’re running AI systems, to validate that the model that AI ran on, at the time that that action was run, had accurate and complete information.
And then you need to have an audit trail. And the beauty of using on-chain systems is it’s fully transparent, fully automated, auditable by everyone. So that brings the transparency, which allows for the accountability for what the AIs are utilizing in terms of data.
Kristin Myers: So let’s talk about agentic AI, because that’s something that everyone is talking about right now, the agents that are actually doing the thing. So if you have an agent that is essentially trying to pull or query your data every couple of seconds...
Rodrigo Coelho: Yeah.
Kristin Myers: Every single day.
Rodrigo Coelho: Yeah.
Kristin Myers: What breaks? Does anything break?
Rodrigo Coelho: Yeah. So that’s been the biggest shift, obviously. What’s changed is that the consumers of data have gone from humans, maybe looking at a dashboard, to now agents needing accurate, fresh data many times a second, right? And so what we’ve built for that is resilient systems. And the interfaces to this data come through what are called MCP servers. So essentially, there’s a trend towards what we’ve called headless software. So you have the data itself, and then you just have an access point for an agent to then access it, which is different than the way a human would. And yeah, these are built to handle machine-level scale.
The other benefit of agentic systems, especially with commerce, which is where we come in with blockchain transactions, is that it allows for micropayments. Traditional financial systems, the credit card systems, for example, with their fee structure, don’t really allow for thousandths of a cent, or, you know, nanopayments. And utilizing blockchain rails and systems like [unclear: “TSA’s”] allows for new use cases that couldn’t exist before blockchains were utilized behind the scenes.
Kristin Myers: So I want to ask about accuracy. I feel like that’s the question everyone always asks whenever AI is involved. What happens when one of those agents pulls data that’s inaccurate, not current, and then they go and run with that?
Rodrigo Coelho: Yeah. So that’s the biggest problem, where we intersect with the blockchain, because there is a nuance with blockchains. You’ve got transactions that are happening over time, and there’s something that happens which is called reorganization, or reorg, where the blockchain has to go backwards in time, maybe a couple of steps, and then go in what’s called a fork and then start a new chain. And this happens regularly.
So imagine that you’ve extracted data from this block, and you have a copy of it. But then the blockchain went back without you knowing, right? And now your data is no longer fresh and accurate. So that’s the big problem with blockchain data. And that’s what our system is meant to handle: we are automatically checking for, updating, and keeping a record of when reorgs happen, and refreshing and updating that data constantly. And that’s really one of the biggest headaches and problems with blockchains.
And when you’re dealing with multiple different blockchains, say you’ve got data from Ethereum and you’re looking at Solana data, each blockchain has a different [unclear: “need”] format and a different way of handling things. So imagine if you’re a financial institution and you’re tracking an asset. Say it’s a stablecoin payment that’s going across jurisdictions, but across chains, or a tokenized stock, like, say, a Tesla stock that was traded on Base, but then went to Solana, but then went to Avalanche, right? You’ve got to track this asset. And so now you’re seeing the problem: you’re taking in all this raw data and then keeping track of where it went, right? But then this blockchain reorganized and reworked, and that way the data is off. So it becomes...
Kristin Myers: My mind is spinning right now. I feel like the Looney Tunes character with the Tweety Bird spinning around.
Rodrigo Coelho: Imagine a giant hairball that you’re trying to untangle, right? That’s what it’s like.
Kristin Myers: So talk to me about AMP and The Graph, because The Graph is decentralized. It’s a decentralized network with a token. Is AMP built the same way, or is that essentially a software business, conventional software?
Rodrigo Coelho: It’s a conventional software business. So when I came in as CEO, that was part of the pivot: building a traditional, sustainable enterprise software business, you know, from our roots in the decentralized world and blockchain world. You know, going from the proverbial hoodies to suits, that’s kind of the change that was made in the organization.
Kristin Myers: Okay. So speaking about that, because just a second ago, my mind was spinning just from the convolution and the complexity that you’re essentially dealing with. We were just talking a couple of minutes ago about how a lot of financial institutions, financial services, are starting to use this technology. These are very traditional businesses. How are you finding those meetings, right, and selling to some of these clients in these more traditional, conventional settings, when your background is in blockchain and in crypto and in Web3, things that they might not necessarily understand? Is that helping you, as you know, or is that hurting you in some ways, trying to bring this sort of new technology and combining it with these very old...
Rodrigo Coelho: Yeah.
Kristin Myers: Bureaucratic systems?
Rodrigo Coelho: Yeah. If you look at job listings across the financial world, there are roles for head of digital assets across the board. And increasingly, they’re recruiting people from the blockchain world to come into the traditional financial world. The reason is they don’t have the requisite background and expertise themselves, and they don’t necessarily want to become blockchain experts themselves, but they need to bring in that talent and knowledge. And so really, having that background is helpful to us.
And we are dealing with, yes, definitely the traditional world. But the people that we are working with are more low-level in terms of the data side of things. They’re, you know, the data engineers, the data scientists, the head of data, for example, that are familiar with data. And at the end of the day, it’s all data, right? And it’s just how we filter it and organize it.
But then you have, on the infrastructure side, they did have to bring in, you know, reliability engineers and software engineers that are familiar with handling blockchain infrastructure. But that’s been a sea change, I would say, starting in kind of 2020, 2021, that there was more recruitment from the... you know, if we go back to the earlier era, it started from the cypherpunk movement. You had people that were interested in cryptography and were kind of the hoodie-wearing developers. And then we have a movement now towards more, you know, serious people with real data. And this data is becoming not just on-chain data. It’s all critical business data that is needed for, you know, really what’s happening in all the financial world.
Kristin Myers: So I have like 15 seconds. I really want to ask you. It’s always our last question. We’re on the floor of the stock exchange. We’ve got the bell right there.
Rodrigo Coelho: Yeah.
Kristin Myers: Do you ever think about taking Edge & Node public? If so, when do you think that might be?
Rodrigo Coelho: Yeah, I certainly see this as a multibillion-dollar opportunity, you know, along the lines of the companies you mentioned. We’re focused on building a viable, sustainable business. Obviously, we’re dealing with the largest institutions in the world. They’re great clients, and they stick around for a long time once we establish a relationship. So I do see a bright future for us.
Kristin Myers: Okay. We’ll have to leave it there. Thank you so much. Rodrigo Coelho, CEO of Edge & Node.
