Alchemy – A Powerful Developer Platform and API for Ethereum Apps

Sebastien: We’re here with Nikila Viswanathan thanks for joining us today. We’ve been wanting to have you on for a little while now. Alchemy has been in the space for somewhere around three years. Recently the company came out of stealth mode after a while. I remember looking into Alchemy in 2018 and it seemed a bit selfie and I remember there was still this, contact us to use the platform thing. Now it’s a full-blown, developer platform that you can create an account. I did that before joining this call and you can play around with the tools and everything. Yeah. Tell us a little bit your background and how you became involved in the blockchain.

Nikil: Yeah, absolutely. Quick background. I actually grew up in a small town in Texas that no one’s ever heard of, there’s no building over four stories tall and everyone drove a pickup truck. That was super fun, came to Stanford, did my undergrad and grad in computer science, machine learning artificial intelligence and had a really great time. Absolutely loved it. Always knew I wanted to do small company startup stuff.

Ironically, over the summers, I ended up doing the exact opposite, which was product management at Google, Facebook and Microsoft, and had started a couple of companies in college. Post-college, I really thought about my life and I realized I went into one of those, 1\8th life crisis thing, and thought about and said, you know what I want to do with my life.

When I think about my life personally, the way I measure it is on two axes, your X axis is how many people’s lives do you touch? And your y-axis is how deeply do you impact their life every day? And it’s basically the number of people, times the depth of impact. For me, that X axis was always obvious, never before in human history, could you press buttons on this magic metal box and build something that every person across the planet used.

It’s so crazy that we live in a day and age with computers, with software, with internet and blockchain, we can build products that affect everyone around the world. This wasn’t possible 20 years ago. When you think about the major shifts that happened in the world, before 20 years ago, the major forces that govern the world, where countries, government, religion, these kinds of things.

Today it’s technology. Technology companies, the Facebook, the Google, the YouTube Instagram have a bigger impact than essentially any country in the world. For me, that was, from a young age, I loved computers. I knew this is what I wanted to do. I had built a bunch of things, at this point in my life probably built 50, 60 different products since I was in sixth grade.

That was just something I love doing. I love creating, I love making things. The Y axes for me were really interesting. We had, how do you really have a daily positive impact on people’s lives? And I had gone through this rough period post college, where I missed all my friends. I was really lonely in Brooklyn, but it wasn’t working out.

I really thought about it. I said, what is the number one thing that determines your happiness on a daily basis? And for me, that was the people I was around. Imagine that time when you’re surrounded by all these incredible, amazing people. You’re just truly happy. I said, I really want to recreate that. If I can do that for every person on the planet, that would be amazing.

Everyone said that’s crazy. It’s so impossible to do, but if you could do it, that would be the best thing ever. We spent several years, my co-founder with Alchemy Joe, who we met at Stanford teaching the database class. It was funny because we TA’ed the database class. Normally it’s an easy class to TA cause you have a hundred students. It’s really fun. It’s really easy that quarter.

We just decided to say, what if we turn it to open it up to anybody in the world can do the class? Turns out 120,000 students signed up for the class. I remember staying up the night before writing code to integrate PDF certificates and all these things. The story is the next quarter that became Coursera. Joe and I were the first unpaid employees of Coursera.

So we met there, we’d been working on products to make you feel like you live with your friends. We built a bunch of different products. It’s really hard. It took a long time and nothing was working. We built this one app. We thought no one would use it. Suddenly it becomes a number one app on the app store on social millions people around the world from page on our times, it’s this crazy story, which I’m happy to go into.

This was around 2015 to 2017. We had seen crypto for a really long time. We actually really believed in it. It was interesting because a couple of the guys who lived in our apartment complex, a couple of units below us were Michael and Jani. They started a company called wire and we used to all hang out and we’d play beer pong and party together.

We have these 3m conversations, how Bitcoin was a future of the world. Joe and I totally believed in it and really excited about it. But I think there was a fundamental shift for us where we had this heart to heart in 2017 and we were wow we’ve seen this happen for a long time, but there’s a massive new shift happening with Ethereum.

Ethereum was really the game changer for us. The reason it was the game changer for us was it was no longer a digital currency, which we totally believed in. But we had said, Hey, we’re going to dedicate our lives to this other thing. But suddenly we realized it was a new paradigm in terms of programming. It was a new fundamental building block.

When you think about the big shifts in technology over the last 50 years, you have the computer, the internet and blockchain and the computer, they fundamentally give you a new building block, each one of these. Computer says, Hey, we’re going to let you create programs. The internet says, we’re going to give you a new building block of instant connectivity. You can connect with anyone anytime, anywhere.

What can you create with that? What blockchain does is it gives you new building blocks you can have, or what Ethereum does, it says you can have programmable money. That was amazing to us. We said, wow this is going to change everything. A lot of our investors had founded a lot of these seminal companies in the internet age or the computer age, whether it’s Yahoo or LinkedIn or Google we’ve seen, we had seen them be in the right place at the right time.

We said, look, this is another massive shift. If blockchain plays out we think it will be, this will be an internet scale technology. We can’t miss this. That’s the Genesis of how we got really excited about it. The rest is history.

Friederike: Talk about your Y axis. Alchemy, it gives us blockchain developing infrastructure. How do you figure blockchain development infrastructure touches people this deeply?

Nikil: That’s a great question. When I think about this, going back to the three big shifts, this is really my mental framework, computer, internet, blockchain, and they actually all formed in a similar way. Let’s just look at the computer. It’s a little hard to visualize without a diagram, which I usually use, but just use your imagination for now.

At the bottom layer, the computer was a bunch of hardware. It was a set of protocols. It was your Ram, your hard drive, your CPU, all the fundamental components that made a computer, but a computer was just a fancy typewriter until you had user applications, right at the very top layer is word Excel, Chrome. These kinds of things that enable computers for everyone around the world to be accessible, besides the people know how to write code.

When you think about that bottom layer, hardware and protocol layer, and then the top layer of applications, the middle layer is the platform layer and what the platform layer does. The computer that was Microsoft windows and Apple Mac iOS. What it does is it abstracts away the hardware and it makes it easy for developers to build applications.

Now what happens is developers can build things so normal users can use those. Then once it has this virtuous cycle, because now users come and they say, Oh wow, there’s a lot of applications. Let me come use this cool thing called the computer. Then more developers come because now there’s more applications and more users, and then they create more applications and so on and so forth. When you look at the internet, you see exactly the same thing.

It’s just a three layer stack. The bottom is HTTP, FTP, SMTP, the protocols that run the internet. There was an internet before the worldwide web, But what is the internet today to us? It’s Facebook, it’s YouTube, it’s Google. That top layer is the application layer. The abstraction layer in the middle that was needed. The developer platform layer is actually the web browser.

That is what abstracts with underlying protocols and makes it easy for developers to build. Similarly in blockchain, what we see is you have Ethereum, you have Bitcoin, you have all these other chains. And that’s the bottom layer. People are trying to build applications, anything from a Coinbase to Walmart doing stuff on the blockchain to crypto kitties to games, whatever it is, these applications can be anything.

In the early days of computers IBM built an operating system, a Deck built an operating system, Altair built their own operating system. Everyone had their own operating system. But then windows came along and said, Hey, we’re going to make it easy for developers. We’re going to standardize this. The way we see ourselves, and obviously each technology is slightly different. The internet is slightly different from computers, we’re slightly different from blockchain.

Fundamentally what we see ourselves as, and what our mission here is to enable the builders, to create applications and provide value to users. If you think about that chain of how blockchain adoption happens, it’s developers build applications and users use those applications and get real value from it. What we’re seeing right now is it’s like the 1991 of the internet, where it’s so difficult to build. It’s so hard to build. Our goal as a company is to enable builders by letting them have the tools they need to build great advocates.

Friederike: And so why focus on the developer infrastructure and more specifically why Ethereum, why this focus on Ethereum when there’s so many other blockchains.

Nikil: Absolutely. That’s a great question. Let me answer the first question first. Why developer infrastructure? At the end of the day, what do we want out of blockchain? What is the purpose of blockchain? The purpose of any new technology? The reason that new technology exists is to provide value to people around the world and enable them to do new things that were never possible before.

This is a sidetrack, but one of my main thoughts around adoption of blockchain is we will know that blockchain is successfully adopted once the word blockchain and crypto is no longer used. Let me explain that. When you go to dinner, you don’t say, Oh, I use this internet app to call a car, to get me here. I’m using a computer app to swipe my credit card and pay, You just say I called an Uber. Then I just use Apple  pay to swipe my card.

Once there’s utility that is provided, that is the technology is irrelevant. We have to get the point where if we’re mentioning blockchain in the product name, then it’s not valuable because it’s not a new technology. We need to get to the point where the technology and the product value from that is the value that the user gets. I’m using this because it’s blockchain.

The second question is, how do we actually get there? How do we get to a state where there’s tremendous value provided by blockchain to people around the world? Which all of us obviously believe in because that’s why we’re here. The answer to that is going back to that chain developers build applications and those users use applications.

The challenge is that the first link in that chain is a critical piece because it’s almost today, imagine trying to build a skyscraper with a hammer and a shovel. That’s the state of the ecosystem today, where it’s so difficult to build applications. Our goal is saying, Hey, instead of having to shovel, if you want to build the empire state building and we give you a crane, we give you a bulldozer. We will give you the power tools and the construction equipment needed to create the applications.

The reason that we chose this, as we said, this is the single most important value add we can do to the ecosystem and help push all ecosystem forward. Then the second question is why Ethereum. The answer is that’s where the majority of developers are, and Alchemy will support and is in the process of supporting any chain that has a lot of developer adoption, because our goal is to push for the entire ecosystem.

We’re not, we don’t take alliances on certain chains. Our goal is saying, people want to create things with blockchain. How do we make that happen? And how do we help the ecosystem and the builders?

Friederike: Let’s talk about what Alchemy actually offers to devs in its Ethereum API. What does that mean as someone who’s never built an app before? What does that mean to me?

Nikil: Absolutely. That’s a great question. Alchemy has a suite of products that we offer. Let me talk about each one really quickly. Fundamentally, anything that’s difficult for developers. We want to make that easy. That is our guiding vision. We want to enable developers to build new applications they couldn’t build before because it was too difficult. We will do whatever it takes to make that happen.

The actual product line. The first challenge we had, so we built it. Joe and I have coded every single day for the last 17 years of our life. All of our team are very similar, very senior engineers who just built a lot of products and worked at a bunch of different companies. We had seen firsthand what tools are in other industries.

When we started building blockchain, we were building other products and it was just so difficult. We had built this hedge fund machine learning, data science platform, that powered assets for, $4 billion in assets for our customers. We were like wow, it is so difficult to build in this space. What we did is we actually ended up talking to about, I think at the time about 80 different customers.

We said, Hey, what are your challenges? Because we have this whole list of challenges. We had built this crazy infrastructure in house that no one else in the world had that enabled us to find out data machine learning that basically was proprietary to us. We said, look, we can just open this up to everyone, but rather than just opening up, how do we actually find out what’s most value to people?

The first challenge that people had was that unanimously that every single, I think out of the 80 people we emailed and asked, I think about probably 77 of them said, this was the number one problem. It was very clear to us. That number one problem was it’s very difficult to run node infrastructure for blockchains. I think also the community people don’t actually realize the Geth and Parity teams have contributed so much.

Ethereum and the community actually owes them a huge debt of gratitude because these people are single-handedly building the tools that power all the things behind the scenes. So I think that was the first thing. We were very impressed and they’re small teams. These are not big teams of these are not, if you think of the number of people who work on Apache, the number of people that work on my SQL, these are orders of magnitudes larger than the teams that work on get them period.

First off they’ve done an incredible job with the resources they have. The second thing is it’s very difficult to run blockchain infrastructure. Let me give you, let me give you a quick example and stop me if this is getting a little bit too technical, but let me give just one example. In traditional Web, let’s say you’re Facebook and you have one web server that is serving traffic for, let’s say you have a hundred users on Facebook.

Suddenly, let’s say Facebook takes off and you have a million users. What do you do? So you actually just spin up more web servers, more computers to him to handle that traffic. Then you split traffic across all of those. The problem is you can’t do that in blockchain. Let’s go through blockchain. The same example, let’s say you’re building crypto kitties and crypto kitties has one node that’s running and node is equivalent to a blockchain server.

Now suddenly imagine you get a bunch of traffic and a bunch of people joining. Now you say, okay, let’s just spin up 10 more nodes. The problem is you can’t actually just split traffic across those. Let me explain why. As a user, what’ll happen, I’ll say, okay, I’m going to buy this crypto kitty. Now that transaction gets routed to node a. But now you hit refresh and say, Oh, I’m going to go view this other cat this time.

You want to see all the cats you own this time, your request gets routed to node B and fundamentally blockchains are decentralized networks and they all don’t have the same information at the same time. Eventually everyone will agree on the same blocks that happened. But at any given point in time, the most recent information is actually not agreed upon by everyone at the network. What happens is the user to the user?

They hit refresh and they’d go look at their cat list and it looks to them like their cat’s not there and their money’s gone. Then they freak out and it’s Whoa, what just happened? Which is not actually the case. The thing is just the transaction hasn’t propagated to the other node in the network. The high level of what that means is you can’t horizontally scale infrastructure in blockchain. That is how it’s done in every other industry.

What we did is we sat down and we said, nodes are great if you’re serving a very light app. We want to be able to scale to enterprise infrastructure and provide scalability. There’s a whole suite of other challenges that people have with nodes. We said, look, we want to make this easy. We actually sat down and built a new type of decentralized infrastructure from scratch. It took years of work for us to build this.

It was a product of us having to build this internally for ourselves first. We call it Supernode. It is a completely new type of decentralized architecture that we have built that enables extreme reliability, infinite scalability. It’s almost like an AWS style thing now where, before, if you had more traffic, you need to spend weeks spinning up a new node depending on the type of node.

If you have a surge of traffic, you’re just hoping you can just spin up a new node on demand. Now you have eight developers who have an AWS like capacity where they can just infinitely scale it will handle all of the data correctness and it handles a bunch of other things around that.

Friederike: Can you become a little bit more technical here? What exactly is this decentralized service that you’ve built? So basically, do you just run your own GETH to open Ethereum nodes in the background and serve that to the person who caused the API? What exactly does your system entail?

Nikil: Absolutely. A great question. We have a diagram on our website. You can check it out, but the short answer is no. Conceptually what we do is we had to build custom decentralized data stores for every specific type of data that is stored in a node. For transactions, you can query transactions in a certain way on the node. We take that data and design a custom data store. That data that allows you to say right now, you can say, Hey, get me a transaction in this block.

One thing in terms of Supernode, what it lets you do is now you can say, give me all the transactions for this address, which is not possible on an actual node. By designing custom data structures for each specific type of data that’s on a node. We basically, we interact with nodes to pull data off, but then we do our own decentralized infrastructure that serves and provides that data to people and provides an additional query layer on top of the Ethereum.

Friederike: Okay, so you built a generalized indexer, et cetera.

Nikil: No, it’s not an indexer in the sense that we’re just storing this in a database and then we provide information. We provide access to it in different ways. It’s truly that. Let me give a good example. Basically think about this. When you’re running, if you want to run a database on your own computer, let’s say you’re building a website. You want to run mySQL and you’re on a computer, that’s a database.

Then Google uses a database to store the entire web and to index the whole Web right? Technically that’s still a database, but the technology is fundamentally very different. Similar here, you can run a node yourself, but the technology that we built is completely different. It’s a new type of indexing and storing and serving method that’s akin to how Google or Facebook would store and serve their data.

Friederike: Can you be a little bit more explicit about how it’s different when you say it’s completely different?

Nikil: Yeah. Let me give, let me give one concrete example. Why is it completely different? The reason it’s completely different is the type of structures that we use to store each type of data is completely different. For example, on a node, you can query Ethereum logs. You can get a certain range of logs.

If you ask for more than that range, it’ll crash, and you can only query it in certain types of patterns. What we did is we said, look, we’re going to use this. For this, for our logs in particular, we use a time series database, and then we also use another indexing method to store the actual information. What that lets us do. We, so we basically store it and serve it in a different way.

We built a whole system just around storing and serving logs. Now on logs, you can search, you can query about, I’m not sure the exact numbers here, but roughly about 10 times the range of what you could query on an Ethereum node, you can query on Alchemy. What it does is it gives developers extended functionality.

Because we built this custom data source for logs, they can query things that they weren’t possible before. Back to the original question Supernode is just one piece of the developer platform. This was the core technology that we worked on for many years to really perfect this and make it really good. On top of that, we have four other products.

Alchemy build is a suite of developer tools that enable debugging, crash, reporting all the tasks that you would do in traditional web development, all the tools that you need, you, those don’t exist in blockchain. We built a suite of tools to enable developers to debug problems faster. For example, one of our users who’s in Hawaii, he had a rogue server that was just spinning and sending a bunch of transactions that he didn’t even realize.

Using these tools, these visualization tools, he was able to find that and shut down, whereas it had been running for weeks and he just didn’t even realize, or another example was there are certain kinds of bugs that take days to debug and with the correct insights, you can S you can debug it in under 15 minutes.

That’s the second piece then there’s Alchemy monitor. What Alchemy monitor does is provide detailed insights and analytics for the blockchain. It captures things that are not possible through a traditional Google analytics or those kinds of tools, because these are blockchain specific. Then the last piece is Alchemy notify. When you think about a normal app, if you use Instagram or use Facebook, you get notifications. When things happen, when someone likes your photo, when someone comments on your wall.

In blockchain, the challenges, these events are not exposed by nodes. For example, when an address receives currency let’s say you send me an Ether, that is not an event that the node sends and says, Hey, this wallet received $5 Ether. What we do is we built a notification system that watches the blockchain for all of these types of events and stores and provides an API.

The end outcome is the user, when you use apps, you can get notifications on important events and that wasn’t possible before. The last thing we do is we actually think of our support as a product. We do real-time global support for all of our customers, 24 seven around the world. It’s not like you’re talking to a support person, you actually talked to the core engineers that built the product. That for us is a really key part of the Alchemy experience.

Sebastien: So just to sum up the product lines, you have the m API, and we’ll get a little bit more into that in a minute. There’s the developer tools that allows the developer to query and this sort of thing. There’s the monitoring tools that allows you to understand your users, data, et cetera. Then there’s this notification thing, which I think is cool where, let’s say you you’re building wallet, you’d like for your users to receive notifications on like blockchain events like on transaction received or on transaction, it’s complete, or some smart contract action or something like that, then that user can receive a notification.

I’d like to spend a little more time on the Supernode thing. Let’s suppose we have someone who’s building Ethereum smart contracts and they’ve spun up their own node. They’re using this one endpoint to build and deploy their DApp. What specifically are the kinds of things that they’re going to be able to do with the Supernode and what kinds of data are they going to be able to pull from that API that they’re not that they’re currently not able to get with the RPC calls that they make to the GEth node?

Nikil: Absolutely. That’s a great question. That breaks down into four key benefits that people, the reason that people use Supernode. The first one is on average, it depends on the size of your project and the size of your team. On average, people spend about a quarter of an engineer’s time maintaining and basically caring for nodes, and typically to build out a reliable node infrastructure system.

Even if it doesn’t have data correctness and these other properties that we talked about, you need three to six months to really make it hard. That is what we’ve seen from a larger customer. The number one reason that people use us is they don’t have to think about it, and it just works, it’s ultra reliable, and it’s just always up. It’s like why you’d use AWS instead of running your own. It just works.

You don’t have to upgrade the hard drive. You don’t have to wait for crashes. It doesn’t go down. If your internet shuts down, you don’t go down. That’s the first thing. The second thing is the scalability that we talked about, the scalability says suddenly, let’s say our traffic doubles, what happens if you’re running your own node, you need to sync a new node from scratch and you need to split traffic over it.

We already talked about the problems with load balancing, but syncing that new node, at minimum, takes many hours. A lot of times it takes weeks. Depending on the type of node you have. That scalability is super important. The other thing is the query APIs. Right now, we have two options. Number one is you can just use the standard Ethereum APIs. We give a little bit of extended functionality on that.

You can query a broader range of logs. You can query trends in certain ways, in an extended capacity that you couldn’t before. Then the second piece of that is we actually provide a suite of higher level API APIs on top of Ethereum that are an extra query layer on top. Let me talk you through a couple of these. One is, let’s say you have an address and you want to get all the tokens in that address, all that tokens that address holds, that’s actually not an Ethereum query.

You have to check every smart contract that you let’s say. Let’s say there’s a thousand, roughly a thousand popular tokens that people might hold. You have to go to every single smart contract for each token and say, Hey, does Sebastian’s address have 0x? No. Okay. Does Sebastian’s address have Augur? No. Okay. Does Sebastian’s address have Kyber? You have to make a thousand quarries just to get the token balance for a single address.

What we did is we said, okay, obviously this is not feasible for any real time wallet or anything like that. We need to build a system where we can, they can give us an address and we’ll give them all the tokens based on the address. So that’s an example of one of the higher level API APIs and like that, the transaction history thing I was talking about, and there’s several other APIs that we power, but the goal here is take a co thing that shouldn’t be complicated and we will make that really easy for developers.

Friederike: How does Alchemy set itself apart from developer products that were out there when you guys started? So I’m thinking of Infura, Fruffle, Alessio, all of these products, how is Alchemy better?

Nikil: Yeah, that’s a good question. I think all of these are fantastic products and I really one of our core philosophies is we don’t really think about other people’s competition because we’re here to expand the space. The market as a whole is very small for crypto. There’s no $10 billion revenue or a hundred billion dollar revenue companies. Our goal really is to help expand the space and we will do whatever we can to help push developer support.

We love it when other people build great tools for this space. This is very collaborative and we are here to expand the ecosystem. I think all of the companies that you mentioned have served really great purposes for different parts of the developer experience. Our core focus at Alchemy is we want to help people. Before we were mostly focused on enterprises and large companies.

Recently, as you guys mentioned, we just opened up to self-serve. Anybody now can sign up and use Alchemy. Our goal is when were building a product, we want to really provide the tools they need to get started. Then also the tools you need to really scale and have a stable surface. For example, most of the tools we provide, we build things when they are not available in other places and when they are available in other places like a truffle, I think that’s a great example.

We don’t build what truffle does. When people ask us, we say, go to truffle, we think they’re great. Our tool set is complimentary in that way. We focus on the areas that we think are not served right now.

Sebastien: Let’s talk about the node infrastructure industry that I find fascinating here, because I think there’s a lot of implications for the broader ecosystem, but maybe some, I think some risks as well. If we look at the ethos of crypto, you talked earlier about AWS and how AWS scales the Web. I think, and a lot of people that work in the space, the way that we should scale crypto, maybe shouldn’t resemble that… we’ll get into that in a second as the ecosystem matures there’s already different layers in this infrastructure stack. So, you have companies like Alchemy that are providing API APIs for Ethereum.

You have other companies in the space are doing more, spin up your own node and they provide you perhaps with some APIs on top as well. Then there’s some other layers more geared towards validation services. I’m talking about companies like Bison Trails, Skills here in France that offer validator as a service or nodes as a service, as the ecosystem matures, you see these types of companies further differentiating themselves or being more and more consolidated. The question there is, do you see Alchemy providing also validation as a service in the future with all the infrastructure that you’ve built? I’m sure it would be fairly easy to set that up for customers. Is that a future towards which you’re tending?

Nikil: That’s a great question. I think all of these different types of services provide really distinct pieces of value. I think at a glance it can look like they’re pretty similar, but when you are running a node as a service, running a validator, running a developer platform, they’re actually very different businesses at their core and they serve very different markets. All the companies you named, we have zero overlap in terms of customer base with any of those companies.

We’ve never had a company say, Oh, we’re decided between you and this other company. So one of the really interesting things, and actually many of those companies that you named actually do compete with each other. We’re just a little bit separate. For us our guiding North star is how do we provide a great developer experience and how do we make it easier for people to build blockchain products? That is what we live, what we eat, what we sleep, what we breathe every single day.

Right now we parking on public stats. We power about 70% of the top of Ethereum applications. We power the majority of DeFi, seven and a half billion a year of on-chain transactions in 99% of countries in the world. The only way we’ve been able to do that is by having laser focus. As of this point in time, developers don’t need validation, staking. That’s not something that developers want in the future. Who knows at least as far as I can see that doesn’t seem to be. Our focus is on developer tooling.

How do we build better analytics? How do we build better debuggers? How do we build better the fundamental construction equipment that developers need and that’s our laser focus? Will we never do it? I don’t think I can say that because I can’t predict the future. If I could, I would definitely like to know what chain’s going to win. But my focus, our focus at Alchemy is just laser-focused on the developer and say anything we can do to make that experience better. That’s what we care about. That’s what we want. As of right now, it seems as far as we can see forward, it’s very much on the same path that we’ve been and continue heads down working on.

Sebastien: Hmm. I mean, do you think it’s possible though? I mean your roadmap, but do you think that, if you look at the way other industries have consolidated, if you look at AWS or Azure and all the types of services that these companies offer within the Web space, as Alchemy grows and becomes hopefully a massive player in the space and a very successful one, do you think that it’s possible for Alchemy to start providing also those other building blocks that make up the infrastructure that powers DeFi? Not just the nodes themselves, but validating services and perhaps other types of services as well.

Nikil: Yeah. Okay. That is a really good question. As we grow, do we expand into those areas? I think fundamentally it boils down to who’s your target, who’s your target market and who’s your customer. The customer is for all the companies that you described are like funds and people who are staking and people who have assets that they want to earn interest on. Right now, our customer is the developer, the engineer, the people who are building these things. A lot of those companies that you mentioned, our indirect competition, are going into the same market and it’s a lucrative market. Happy for them.

For us, or our core focus is developer. As of the current present moment, we don’t see a developer, a core need of the developers is not necessarily that they need to stake. Maybe for their own personal assets, but that’s not a core thing that they need right now. If that is the case, let’s say crypto evolves in a way that developers need mechanisms to be able to stake assets in order to build things, share. Maybe we’ll support it then.

The thing I can very confidently say right now is we are not trying to build tools for a hedge fund to stake their assets. That’s not our core business, our core business is whatever developers need. In the sense of developers in the future, crypto pivoting and enabling developers to do that, we would definitely support that.

Friederike: Okay. Currently you a subscription service. Basically I pay a flat fee and then I get to make a number of API calls. What do you currently do with the data trail that the user leaves? If I connect to your service and they make an API call, you can make the connection between my IP address and my contract keys I hold and the interactions I engage in, which is potentially extremely valuable information. Do you data mine that? Do you save that? What’s your, what’s your commitment to using the data?

Nikil: So one of the really nice things of our business, which I’m very glad about it, cause we had, you know the legal diligence done on us. Fundamentally we don’t store any private keys. People have asked us a lot, can you optimize my transactions? And can you store my private keys? We’re Nope, we’re not going to touch that because that’s a whole another issue when we start controlling users’ funds.

In terms of funds, we have no access. Let me give you a quick mental model of how you can think about this. We are a pipe to the blockchain. There is a blockchain here and there’s a user here and the user needs to take a packaged piece of data and put it on the blockchain. We’re like a good, fast, reliable pipe, but ultimately we’re a pipe to the blockchain.

Friederike: I see. But your pipe that knows what traffic goes through it. I never thought that you guys kept private keys or whatever. But basically you are able to make the connection between my IP address and the information that I carry and the transactions I send that is potentially super valuable, no?

Nikil: Yup. That’s a great question. The question is do we have an IP address and this package of data. Two parts to that. The first part is the transactions are all public. There’s no secret information there. In terms of the IP address, that is the only piece of data that we do have. Your IP address can be spoofed. It can be fake. We don’t even know if it’s a real IP address. We take user data very seriously. Even though no one’s complained or no one’s asked about this. We don’t sell the data.

Our business is not, we’ve had a lot of people ask can you sell data to us? That’s just not our business. Our business is providing customers with a great developer experience. We do nothing with the data. We have systems. We can’t even look at the data ourselves and to be clear, we don’t actually have any data. The only thing we have is an IP address of the user. That’s all we have. We don’t have their name. We don’t have their email. We don’t have anything else.

Sebastien: But you might have, you might have transaction data. I suppose with the monitoring tools that you do have, I mean, there must be some data about transactions that’s available for developers to use. For example, what transaction or what data are my clients or users of my client, basically what API calls are being made and what transaction requests are being made into which smart contracts, et cetera. This metadata, not necessarily the direct user data like names and email addresses, but this metadata must exist somewhere on your app, on your systems. Does it not?

Nikil: Yeah. We know which API calls are made, but again there’s no user identifiable information. We don’t store it. We let the user debug their product, but we don’t look at it. We don’t dive into it. We definitely don’t sell it. Not that, I know there are companies that do that and nothing against that. We just chose for our core business to be providing the service. It’s really interesting because there are certain types of data that people are very sensitive about, whether it’s private keys or identity, if we could identify which users had addresses or which users were sending transactions, or which users were doing this, that would be a whole nother story.

Friederike: But Nicola. In fact, you can. Basically if you look at how easy it is to actually identify addresses on the blockchain, if you now also have the IP data that goes with it becomes even easier. I totally agree that IP data can be spoofed and so on, but being diligent about this 100% of the time is nigh impossible. That’s the reason why whistleblowers have such a hard life, because basically staying truly anonymous on the internet is really hard.

Nikil: That’s true. I think there are two points. One is, again, literally the only piece of data we have is IP, even for our own internal metrics. When we think about how many users use your platform, it’s actually really hard to tell because IPs are not correlated to actual users very clearly. Your phone. If you drive another mile on your phone picks up a new IP address. It’s very unmatchable to people.

The second thing is most of our traffic, actually, the majority of our traffic is backend services. Most of the IPS we have are just from Amazon. People’s own AWS clusters hitting us. I don’t know the exact stat off the top of my head, but I know it’s about 70% of that is actually backend traffic from customers.

Friederike: As a user of dApps, as someone whose privacy consent, forgive me, I’m driving. As someone who is privacy conscious, I want to know if a DApp uses a service like yours, what are your thoughts as to this? Because basically if I connect to DApp and in reality, actually connect to you, that’s something that the user should know, right?

Nikil: When you have a DApp you’re connecting either to that company, if they run their own nodes or you’re connecting to upbeat. Diving into the technical architecture of a DApp here, there’s two ways a DApp can be constructed. Either the company can say we’re going to route all the transactions through our own proprietary systems. Then that backend system queries Alchemy for data and returns it. In that case, we get zero information about, we don’t even get the IP address.

They can use the Web Three on the front end, which means the query’s are made from that user’s computer. In that case, the user, the only information that’s passed is the IP address. In either case, and one thing that’s interesting, and this goes to the question about centralization decentralization.

One thing that’s interesting is when you’re using a DApp, if the company runs the actual software and the node stuff and the nodes, it’s essentially like you’re using a centralized service, that’s hitting that individual company. But with Alchemy, the front end is served by the product and then the backend is served by Alchemy. It’s almost like in a way you’re, you’re only depending, partially on that individual company.

Sebastien: I’d like to turn it back to the like blockchain APIs. Do you have an idea today of how many nodes, if you look at the entire Ethereum Network, do you have any idea of how many dApps in the broader data ecosystem use Alchemy, and perhaps also, one of your main competitors, Infura as their backend.

Nikil: Great question. What can I say publicly? We have thousands of developers using us, so we have a lot. We have it got to the point where it’s actually pretty cool. Over the last couple of years, in the beginning we would get so excited whenever we got a customer. I mean, we still get very excited when we get a customer. Ultimately that’s what we’re here for is to help people build great applications.

Now it’s got to the point where the majority of time I used crypto stuff, I would say a huge percentage of the time I find out I can like look into our look and say Oh, we have customer support chats with all of our customers and telegram, and I’ll just search that company’s name. We have a customer support chat with them. Oh cool. They’re using Alchemy.

One of the really cool things is, as we’ve seen in terms of the growth of the ecosystem we’ve seen as DeFi’s been blowing up, we’ve seen a ton of really cool things. I was explaining this to a friend and I think one of the most interesting things about the DeFi ecosystem, what I think fueled the growth of the DeFi ecosystem, is every DeFi project effectively has an open read-write API for anybody to use.

Imagine if like Facebook, Google, Amazon, they all had read-write API that anybody could build on. It would be more difficult for Facebook, But the innovation that would happen on top of that would be mind blowing. I think that’s one of the really cool things that you’ve seen in terms of speeding up DeFi and for us, in terms of the growth of our user base, it’s been really cool to see that because we directly correlate to the number of projects in the space.

We have a certain percentage and with that, it grows. That was the main reason for us spending months and months building out self-serve where anybody can sign up, can use it, which is actually very tricky infrastructure and developer challenge to solve. It took us probably about five months to build this and our whole focus there was before we only served the largest companies.

Now we said, we want to make blockchain accessible to everybody. That was always the goal, but it’s very difficult technically to do. Now, we start seeing people on our free tier who are very tiny companies and then suddenly they’ll blow up and there’ll be one of the largest people in the space. They started out from that small free tier.

Sebastien: I want to come back to this topic of AWS, which to a lot of people in this space, scaling Ethereum and scaling the blockchain translates to increasing decentralization essentially. Having more nodes, having a diverse validator set or mining power, what do you say to developers or people in the ecosystem who fear that services like Alchemy and other node aggregation services or APIs where essentially there’s aggregation of, or centralization of access. What do you say to those people who think that this is detrimental or somehow goes against this goal of bringing further decentralization? I want to give you a concrete example. If in five or 10 years from now 80% or more of dApps on Ethereum or are using one or two API, do you think that’s a success for Ethereum? Do you see that as a success for Ethereum or do you see that as a risk potentially for powerful entities governments, for example, to step in and break DeFi?

Nikil: Yeah, absolutely. I think that’s a great question and something we thought about a lot. In the very beginning, before we built anything, we were convinced that we needed to build a pure decentralized solution for this. That’s what we actually started building. We actually built all the tech. It’s actually a very complicated technology problem to do a decentralized network to support this. Take something like auger, we have a betting market, something happens and you need to settle the results of a real world event and you need to settle it within a couple of days or whatever. There’s a real world event. Someone sends someone up on a piece of data and we need to settle that in like 20 milliseconds to know if that’s correct or not.

There’s this thing where, just having someone stake and then slash if they get wrong, isn’t actually going to work because the damage has been done, that incorrect data has been sent. That could have massive consequences. It’s this very complicated structure that we spent months thinking about. We had some of the math Olympiad type minds working on it to figure out a really cool solution for it.

We had, we actually ended up coming up with a really good solution. Then we ended up talking to about 80 customers and we just asked them how much do you care about this? How much do you want a decentralized product? How much do you want? Whatever. And, honestly, we were blown away and this is one of the things about user feedback, where you do something and you think it’s so obvious and everyone knows.

Then users are Oh, actually you’re an idiot. The people who designed it, we just said, we just didn’t realize. What we found was, number one, people didn’t care. They said, we want something that works. Because when you think about this, there’s two functions. When you think about it, the first stage is getting a product that works. Then it’s like having a completely, truly decentralized product. Right now, when you think, and then the second beat, and everyone’s just focused on that first piece now making something that works.

The second thing that’s super important to understand, and this was the crux of the reason that people said, we want this solution first. We actually have the decentralized version in our back pocket. When the time is ready and when we feel like we need it, people will, we’ll build it out.

The thing is, there’s a very big difference between us and AWS. In terms of the centralization aspect and the answer, it goes back to that thing where we’re a pipe to the blockchain. We don’t store any blockchain data. The data we give, you can get from anywhere else. Which is why the Supernode is actually a small part of our business. The developer tools are a huge chunk of our business. The reason for that is people can switch off at any time. Essentially, one of the things we provide is we provide a commodity service. Of course, we do it better, more reliably.

Sebastien: There will always be the alternative. This is a sort of criticism, not on Alchemy, just in terms of the vision for what blockchain systems should look like. Do you think that it would be a success if, 80% or more, or it’s like 90% or some huge number of dApps were pointing at two API URLs?

Basically Two API endpoints or one API endpoint. Would you see that as potentially a risk for DeFi to get shut down nodes even if your servers go down or if your competitors servers go down and all of a sudden there’s 50% of the ecosystem that goes down with it and DeFi, by this time, the entire economy goes down. Well, the DeFi economy, what does that catastrophic scenario look like in your view? Is that desirable given where we know blockchain comes from, it comes from this desire to have things that are decentralized that are resilient, et cetera.

Nikil: Absolutely. I think for us to be a success in blockchain means that we’re providing new value and new products that we couldn’t do before, and that it cannot be shut out. The not being shut down is a really important piece. I would say if Alchemy was more like an Amazon service, Amazon has all these companies data, and if Amazon goes down they’re not getting their data back.

I think a success case for us in Alchemy would be actually that, nodes just get a lot better and they, everybody can run their own node and that works and we’re just focusing on building the tooling around it. Our goal is not to be, we don’t want to be a node infrastructure company. It’s actually not a great business to be in the developer tools, it’s actually a much better business from the business side, because nodes are very expensive to run.

When we think about the DeFi case, let’s say there’s two API APIs providing or one API providing, the majority. Right now we do about 70% of the top apps. But let’s say that, I think there’s two cases here. There’s a case where, if God forbid, something happened and the government comes and says, Hey, we’ve got to shut this down. If the case was that these companies would be hosed, that would not be good. That would be a very bad thing. But if the case was, the company could just run their node point.

One of the nice things about Alchemy, which is a great thing for the customer, not so great for us, but you can switch on and off Alchemy and one line of code literally all you do is you take your line of code and you point at your own server. It takes less than three minutes to switch on less than three minutes to switch off, which is tough because what that means is we have to continually win our customer’s business, because it could be anybody just put it off at any time.

I think the way that it’s going right now is we’re providing a suite of tools on top of node functionality, which I think are very valuable, but people are not dependent on us as much as for our business. Our investors would love to have people depending on this, but the only way that we’re able to do this is by providing continually good service.

Sebastien: I get what you’re saying. I get the argument that they can just change the URL\endpoint. Then they can point somewhere else.

Nikil: And all their data is still there. Everything works out.

Sebastien: Of course all the data is still there. I think it might be a bit harder, because you also offer these additional features, there’s some lock-in there and they might have to rejigger their product without those additional services and those additional API endpoints. I’m just thinking more in terms of ecosystem resiliency.

What I would hope is that the ecosystem is spread out in such a way that, there’s 20 Alchemy. There’s several companies like yours offering varying types of services and maybe even niche services for certain types of dApps that provide them specific types of API calls. That we have a more resilient ecosystem that doesn’t rely on very large actors that can easily take down the entire space.

Nikil: Absolutely. I think, again, I think that’s super important. In realistically, what would happen if a company switches off Alchemy, the transition would be easy, but continuing development on your product would just be slower. I think the trade off here is we thought about this internally, do we build a decentralized system? The question for us was, let’s say we can push the space two years ahead by building the centralized version first, because it’s easier to build and we’re able to provide better tooling and better developer experience on top of it versus taking two years longer to build a decentralized version.

For us, when we think about this, because nothing in life is free. When you think about building a decentralized version of product versus developer tooling, that is more of a centralized pipe. The trade-off there is the time to market and how fast you accelerate the ecosystem. So if we felt like it was a case where people would be stuck on us and you couldn’t switch off and it’d be shut down, then we wouldn’t have done it.

In this case, we said, Hey, we can push the ecosystem two years ahead by providing developers, the tools they need to build things, in one fifth the time, and they can switch off at any time. I think that’s the trade off you have to make. We fully believe that there should not be any single point of failure or even a few points of failure in the ecosystem. Right now we don’t think that’s the case.

Friederike: Maybe let’s switch gears and get to the last section. I’d like to talk about your investors. You actually have a very eclectic suite of investors, and Sebastien made me not start with this, but how did you meet Jay-Z?

Nikil: It’s actually really funny that I get that question. We get that question a lot. First off, I think, let me preface this with those stories. The first time that Joe and I, my co-founder and I started working at it, literally the very first day, we were walking up the stairs, were in the Stanford incubator and I asked what music do you like? His girlfriend, Sarah, was there. Sarah says he loves teenage girl pop. My answer was perfect. That’s exactly what I love. We’re going to get along super well.

I have actually, never really listened to Jay-Z’s music. I’ve never really had play on it. I don’t really listen to rap. Jay-Z and a bunch of other investors have been inbound at us. We were always, look, we just want to build a product. We’ve been in a very fortunate position where we ran our company super lean and we never needed cash. It was always inbound interest. One investor was Hey, Jay-Z really wants to meet you guys.

We’re sorry. We’re heads down working. During this time the company is going really crazy and we’re working really hard. There were days we hadn’t left our parking lot six days a row. So, he got this call, Jay-Z really wants to meet you. We’re ah, we really got to focus. No, he really wants to talk to you. I was like all right. We had just wrapped up doing a fundraising round. I just honestly didn’t want to do any more investor meetings. I wanted to get back to work.

We said, look, okay. The way relationships we have with our investors is we really want coaches and mentors. We didn’t need cash, but we said, look, we want personal access. We want texts. We want to call you, email. We want to learn. Whether it’s Reed Hoffman or Charles Schwab or the tremor board of Google, we’ve been very lucky to learn from all these incredible people. They said, Hey, you don’t have any media entertainment people, and they’re really smart. You should meet them.

I’m all right, cool, let’s jump on the phone tomorrow. They, and then they said, Oh, I can’t because Beyonce had just launched her album that day and they’re going on tour. I was all right, well I guess it’s fine. It’s not gonna work out. It’s totally fine. We got this call from Jay Z the next day. He, and he calls us, he’s Oh, I had to eat my dinner really fast to talk to you guys. We ended up having a really great conversation. We interviewed him, asked about his life and what he was looking for, why he wanted to work with us in these things. Was he excited to mentor and coach us?

And honestly, I have to say I was blown away. I listened to One Direction and Taylor Swift. I literally knew nothing about Jay Z and he’s a super smart guy, very kind, very charismatic. Yeah, we really liked him and all right, welcome to the family.

Friederike: So what are Jay-Z’s thoughts on Ethereum?

Nikil: We talk more about the business side, around how to run a business, how to think about your team and brand and stuff. Then there’s other things that we talk to about, we don’t necessarily need every investor to be deep into developer tooling.

Friederike: We’re also wondering about this before we got on the show, is EULA Stanford one of your investors? We didn’t know that Stanford had a VC arm or an investment arm. What is that about?

Nikil: A couple of different venture arms out of Stanford. There’s a SanFran endowment. There is the Stanford StartX fund. There’s The President’s Fund. Sanford directly invested and we have a really close relationship with them. This was also when John Hennessy, the tremor board of Google was present at the time. He also personally invested, so we’ve been very tight, and in the CS department, multiple of the computer science professors are investors. Just really close. I spent six years there, undergrad and grad, Joe did the same, so a lot of love for our school.