Moderator: Blair Zhu, Mint Ventures

Guest: Ganesh Swami, CEO of Covalent

Interview time: March 23, 2024

Original interview link: WEB3 Founders Real Talk with Covalent Recap



Ganesh’s experience and introduction to Covalent

Blair: Hello everyone, welcome back to Web3 Founders Real Talk. Here we will connect with real industry disruptors and have honest conversations. Today, we are very happy to invite Ganesh, CEO of Covalent. Welcome!

Ganesh: Blair, I’m so glad to have me on the show and I look forward to further interacting with the community.

Blair: Thank you very much for coming. Can you briefly introduce yourself? How did you get into the cryptocurrency industry and how did you start your project? Also, please tell us briefly about Covalent.

Ganesh: I am Ganesh Swami, one of the founders of Covalent. Covalent has been established for more than five years and is considered one of the established projects. My foray into crypto was rather serendipitous. It just so happened that I got an opportunity to enter this field. I didn't work in the crypto space before, I actually came from the database space, building data infrastructure. Before that, I worked in cancer research. I did physical chemistry research and built antibodies for drug design. I am a founding team member of one of Canada's largest biotech companies, which is listed on NASDAQ and has several drugs in clinical trials. This is my background. I switched careers because pharma takes 10 years to build a minimum viable product, it takes time. My friends in the IT industry only need 2 years to launch MVP, enter the market, raise funds, and conduct mergers and acquisitions. I wanted that fast pace too. So I turned to data infrastructure. At that time cloud data warehouses became popular, such as Snowflake. Many local work tasks have been moved to the cloud. The cloud is like a new piece of infrastructure. I've helped many companies move to the cloud. I worked in a coworking space for about ten years. One of my mentors told me that you should go to a decentralized database project hackathon. That was during the bull market of 2017. I was like, okay, I'm in Vancouver, and it rains a lot in Vancouver, so I don't have anything else to do on Saturday, so just go check it out. I know in the database world, ultimately it doesn't matter what database you use, people still want to do their analysis in Excel. That's the front end of all databases. Whether it's Oracle, SAP or Microsoft, it's the same. What I developed during that hackathon was a way to import blockchain transactions directly into Excel. That's the idea. It’s like a search engine, Google on the blockchain, whatever you want to call it. That was in 2017, there were only ICOs, it was simple like ERC20 and transfers. That's all. No DeFi, no NFTs, nothing complicated. We ended up winning that hackathon and we said, this is a cool idea that could start a lot of things.But what was I wrong about? It was market timing. Because the next two years were a bear market, and it was brutal. So please don’t take my advice on market timing, I have a terrible track record. Anyway, we started this company called Covalent. If you remember your high school chemistry class, the word Covalent comes from the word covalent bond in the chemistry vocabulary. We tied centralized systems and decentralized systems, databases and blockchains, and so on. That’s an analogy. That’s the origin story of Covalent. We started this company, which was essentially a blockchain as a database. You could query it, index it, do all kinds of operations from the blockchain. The first few years were a rough time. Then DeFi Summer came along, and the timing was right, and the product was right. People will say it was like an overnight success, but we had been working on this for about two and a half years at that point. There were some ups and downs along the way, but overall, this is our startup story. If it hadn’t rained in Vancouver that Saturday, there might not have been Covalent. It’s that simple. The core idea here is that no matter what infrastructure changes happened behind the scenes, people don’t relearn or change tools. They don’t give up Excel. They don’t retrain existing workflows, and existing business processes have to adapt to the changes. So people need this bridge broker, like a middleman. That's what Covalent has been providing since day one. We've never really pivoted or anything, that's our model. Today, it's been given different names, some call it an indexer, some call it a data availability layer, and so on. But fundamentally, what we do is make blockchain data more accessible in a decentralized way.

Challenges and resistance

Blair: That's the most interesting story I've ever heard. It's all about the weather, and I'm glad it rains a lot in Vancouver, which is why you were able to start Covalent. You found an important product pain point and started this amazing project, and it's doing great now. I'm wondering if you've faced any challenges or resistance along the way? Like you mentioned, the timing in the context of the entire crypto industry, that's probably a key indicator. Did you run into any technical challenges or macro challenges or anything like that?

Ganesh: I am a serial entrepreneur and this is my fourth entrepreneurial project. There are risks associated with any mission-driven entrepreneurial project, and these risks are often bundled together. For example, there is market risk. This is where we failed in the first two or three years. There are also product risks, technology risks, financing risks and team risks. All these risks are intertwined. We faced financing risks because no one was willing to pay in a bear market. There was also market risk, because the market didn't exist at that time and there were no applications. We built the product. We're pretty good engineers, a good team, and my other co-founder, Levi, has been building databases his whole life. There is no doubt that he knows a lot more than I do. Of course, there will also be financing risks and technical risks. On the other hand, we were lucky because EVM won out and basically everything became EVM. There are teams betting on EOS, Cardano, and Solana, and they are doing very well. But any projects that bet on non-EVM, such as XRP and Elrond, have now failed. Technically, we are very lucky to choose EVM. These are some of the risks, but the biggest cruelty is financing and market risk. After two years of working day in and day out at Covalent, we did not rely on outside capital. I was disillusioned with this business. It wasn't just me who was disappointed, a lot of people exited in the bear market, just like you saw in previous bear markets. One of my mentors suggested that I take some time off to get a different perspective and see if this industry was really for me. So I climbed Mount Everest. It was a very difficult journey, but I spent a lot of time alone, eight, nine, ten hours walking alone. The whole team was together, but I was immersed in my own thoughts and had a lot of time to think about things. I gained some great insights in the Himalayas. One of them is, now that we've crawled all this blockchain data, why not leverage it, see its traction and do outreach to see if others want our product? So we came back.I came back in October, November, December, January and February, and I made it through those four months, including the Christmas holidays. We found product-market fit, started to make profits, and gained consensus. This is where the flywheel begins. Another chance is like a different perspective. We have all the data and can see the share of different protocols. So why not contact them? I would say these are the huge headwinds and challenges. Then there's also the challenge that people don't understand the value of indexers because all the data is public. What is the difference between Etherscan and Covalent? What is the difference between The Graph and Covalent? But these are ongoing, it’s part of the journey. But first I would say the biggest challenge was getting this work off the ground, and it took us almost three years to get some glimmer of hope. Those days were long and hard.

Blair: Yes, but it’s also impressive because the space is still so new and we still see some entrepreneurs struggling with product-market fit. Sometimes I think entrepreneurs need to be selective about what they've been doing, not just because of their interest or a certain opinion.

Ganesh: There is a thing called product founder market fit.

How it differs from other data solutions

Blair: Yeah, that's exactly what I wanted to talk about. Can you give us a quick overview of your product range? Because I see you have a Unified API and GoldRush. Also, what cost savings can developers expect with your product compared to manually handling data retrieval and processing through RPC? I'm not from a technical background, but I think this might be a question that some people are wondering about, what the difference is between the two. Also, how is it different from other blockchain data solutions like The Graph?

Ganesh: Okay, that's a great question. Maybe before we talk about different kinds of data solutions, the key thing is that blockchains are billboards, not databases. On a billboard, you post something, and a week later, you take it down and post something new. That's the blockchain. The whole purpose of a blockchain is to put it into a challenge window and see if there are any challenges. After the challenge, you evict it, expel it, and then move on to the next operation. You evolve the state machine, which is the core of the blockchain. A lot of people misunderstand this. They don't understand that blockchains are for state propagation, not for storing any historical data. That's a key point. The second problem is that every blockchain has nuances, some use POS, some use POW. You see all kinds of new Rollups and DA solutions. Some use Call Data, some use Blob storage. From a developer's perspective, they just want to see token balances, NFTs, cost basis, and standard information. They don't care about the technical details, it's not important. So a unified approach makes sense. This is very new for Covalent, we built a unified interface for all the blockchains we index. We index about 200 blockchains, including testnets. So if you integrate Ethereum, you just change one character, and then you can integrate any other blockchain, like Polygon, Arbitrum, Phantom, Optimism, Base, Mantle, and so on. For any EVM chain, you just change one character. You build the user interface, and then everything just works. This is very popular among developers because they don't want to rebuild the entire stack over and over again for all the different chains. There are also some important aspects, just understanding the data stack itself. You have data products like Coingecko, which are more for retail users, so they provide high-level statistics, market capitalization, circulating supply, etc. This is not really on-chain data because some of this data is also off-chain, but it is provided for retail users. Then there are infrastructure layer indexers, like Covalent and The Graph, which provide structured data.Then there is RPC, which is a lower layer, such as Alchemy and QuickNode providing raw data. That's the level. Our expertise lies in structured data, as RPC gives unstructured and messy data. This is one of the key differences. The value of an indexer is in taking all this raw, unstructured data from RPCs or the blockchain and presenting it in a usable, consumable and readable structured way. This is how the entire stack is organized. Moving on to The Graph, I think there are two different philosophies when it comes to building indexers. The Graph has subgraphs, while Covalent has a unified API. In a subgraph, you create DApp-specific endpoints, but each DApp has its own schema and structure. And in Covalent's way, it's a unified pattern. It's not specific to a DAPP or anything else. The use cases, traction, and customer segments are all completely different, but they all solve the same problem to some degree. What's interesting is that about five years later, The Graph became more like Covalent, and Covalent became more like The Graph, because everyone was trying to expand their scope.

The process of creating the flywheel effect

Blair: That's interesting. I noticed that you guys highlighted a lot of things to achieve in Covalent Vision 2024. Now that we're a quarter in, you mentioned that EWM is very critical, which I understand. But I wanted to ask how your team makes decisions on the roadmap? Because there are so many things involved. Can you give us an overview of where we are right now? Which of these projects will be the focus of your attention?

Ganesh: I think there is some order in this mess. Although it seems like a lot of things, it is actually like a puzzle, a holistic project. Let’s take a step back and understand this flywheel. Covalent indexes the blockchain. We index the blockchain, so developers and DApps on those blockchains use Covalent. After they use the product, these DApps want to implement multi-chain operations. So they get more data from Covalent, which means more use cases are unlocked. As more developers and use cases are unlocked, more blockchains will want to join and take advantage of those use cases and developers. We've indexed 50, 60, 70 blockchains, and all of them have been attracted. We do not engage in any active promotion. They want to say, come on over, they have all these products and attention. Take, for example, Rainbow Wallet. Rainbow wallets are very popular. All data comes from Covalent. They won't go to the new chain unless Covalent supports it. Therefore, we receive a lot of requests from DApps. For example, we announced the Blast index, not because the Blast team asked us for it or we reached out to the Blast team, but because Rainbow approached us about supporting Blast. For Rainbow, only one character needs to be changed. They only had to change one character and suddenly Blast was supported. Everything is supported on Blast. So they don't need to rebuild anything. This is very convenient to use. This is the flywheel effect. All of this is like a constantly spinning flywheel. The key is the involvement of tokens. All revenue from the demand side, from developers, is a pay-as-you-go API, so it starts out free and then starts paying out revenue. These revenues flow to the operators of the operating nodes, who are also CQT holders. This is how the entire flywheel rotates, and this is how the decentralized economy began to develop. This may seem huge, but there’s so much more to come about our community programs, fee buybacks and conversions, EWM, our product list on the demand side, our developer grant program, all of our Rollups and Rollup as a Service Indexes are all part of this giant flywheel.It will continue to spin faster and faster. It's all part of the same plan, it just seems like discrete components.

Product development progress

Blair: It seems like things are a little disjointed, but like you said, everything is connected. How is it going now? What kind of product development can we expect to see in the future? Can you tell us a little bit about that?

Ganesh: No problem. The key point here is one of the points that we highlighted in the review last year, which is the fee conversion mechanism, which is the proceeds from external sources, which is basically the proceeds from customers paying the proceeds, which is used to buy back CQT and distribute it to the operators. This mechanism started about 45 days ago, and it bought $1,000 worth of CQT every day. Maybe in the show script or somewhere else, I can share the wallet address. It bought $1,000 every day. Sometimes the price of CQT is 20 cents, sometimes it's 40 cents. It doesn't matter. As the demand side proceeds increase, it will basically determine the price floor of CQT because it's buying anybody who wants to sell. That's the price floor. This mechanism is now live. This is an exciting update and the final picture. The other thing is the staking migration, back to Ethereum. So far, we have been using Moonbeam for settlement and so on. I think in general the Polkadot ecosystem is not as good as it should be. So we migrated the staking back to Ethereum. All the audits have been completed. The next thing is the EWM testnet, the incentivized testnet. It's almost ready. Then double down on the AI ​​and DA narrative, build more products and engage in the community. Everything is going according to plan.

Use cases for AI models

Blair: Hope everything goes well, it sounds like a lot of work. I learned from your social media that Covalent is making a big push into AI and can provide very rich historical and real-time Web3 data sets. How does this work? Can you cite some specific use cases of AI models? Web3 and AI have been interoperating for a while, but we're currently discussing some really legitimate use cases. Can you name some?

Ganesh: No problem. The key to large language models, or LLMs, is structured data. This is the input for everything. This is the amount of data required to train these large language models. The whole point of Covalent is to have all this structured data. You can feed all this structured data into these language models, and then you can fine-tune the existing base model, any model you want, and then you can start making inferences on it. That's the whole process. This is very similar to getting structured data, then running a query node, and then querying that structured data. This is a database product. You just transition from big data to big models. It's like a transition. It was a very natural extension for us. We were quite surprised when the market started adopting Covalent for these use cases. This makes sense. Structured data is like cleanly formatted, normalized data. Who doesn't want that? So we're starting to see a lot of use cases. Recently we published a post about all the AI ​​use cases in construction today. Maybe we can put that in our show notes or something. Smart Wheels are one example. Smart Wheels is a platform for on-chain copy transactions. You can follow any type of wallet and they do a summary of multiple wallets and then they use AI to determine if this is a trap or a scam. Smart Wheels is a great example of a project doing very interesting things. Another example is Leica. Leica.AI uses AI for analysis. You can see many examples here. Again, we're not on the analytics level, and we're not on the retail side. They use all this data for training and can then perform sophisticated analysis on the tokens if you want to do research or other uses. Leica is a great product for this. Another cool thing I heard about recently is Entendre Finance. It provides anomaly detection and predictive analytics, which is very attractive for financial management. In the background, they look at your payroll and expenses and more. They can leverage AI for fraud detection. Another example is bitsCrunch. bitsCrunch is a project that recently went public. They conducted a coinless sale.Their investors include Animoca and Coinbase. They utilize Covalent data for various operations including fraud analysis. Therefore, the basic data behind these projects are provided by Covalent. These are just some use cases, like we started Covalent before DeFi, NFTs, and GameFi existed. The market develops in different ways, but that's at the application layer. We are at the infrastructure layer and we can enable all of these use cases.

Application of CQT in the ecosystem

Blair: That's impressive. It's great to see you empowering these innovations and that people are sparking change and making an impact because of Covalent. You mentioned CQT several times in today’s conversation. Can you further elaborate on the specific role the token plays in the ecosystem? I feel like this might be another unique thing compared to other products. Additionally, you recently launched a token buyback program to turn offline revenue into on-chain revenue. Can you share more insights on this?

Ganesh: CQT stands for Covalent Query Token, which is a token used for staking and governance, and it’s very important to the Covalent token economy. From our experience in the market, developers and consumers of products don’t want to use tokens for payment. It’s kind of like a bifurcation. The tokens in the hands of these people are like antiques from 2017, and they don’t make sense. So everything on the demand side, all the revenue that customers are charged is denominated in US dollars. It’s fixed. There are no challenges, forecast budgets or anything else. So the US dollars are then used as an on-chain mechanism to buy CQT. When I mentioned a thousand dollars a day, that thousand dollars is from customers. And then the CQT purchased based on the market price is distributed to the decentralized operators who actually perform the work. They get CQT. So the analogy is, I hired some contractors in the Philippines and I paid them in their local currency because that’s the currency they use to spend. I can pay them in US dollars, but they will convert it to their local currency. That’s it, the entire Covalent economy is based on CQT. There are other utility functions, like we will launch a program like liquidity staking, and participating in the delegated staking program is another utility function of CQT. As a token holder, you can choose to delegate your CQT to one of the operators. We have about 14 operators, and we're going to publish a post to recruit more operators. They're responsible for running the actual infrastructure. You can delegate your tokens to them. We have a whole token economics program around incentives for long-term data availability, and even use cases for AI. Maybe you saw the New York Times lawsuit, where they sued OpenAI because they used all the New York Times articles to train their data. So if there is any bias or any gains, then the royalties have to flow back, which means you need to record all the mutations that are made to the base model. All of this, like blockchain, is a perfect use case for this kind of thing. Going back to the token, it's a normal ERC20 token. It can be traded on OKEx, Uniswap, Sushiswap, KuCoin, and Gate.You can hold this token and participate in this economic system. As a principal, you can hold CQT and earn income. Or, if you have enough technology to run the infrastructure, you can also become an operator. This is roughly how the system works behind the scenes. In addition, you can also provide LP in DeFi.

Strategic plan for revenue growth

Blair: It's a very well-designed mechanic, especially for all the stakeholders in the game, who are all incentivized. Can you outline your strategic plan given your very ambitious revenue growth goals? Given your institution's steady growth in users, do you foresee a unified API significantly driving this growth? You now have two pillars in your product line, one is GoldRush and the other is Unified API. Which one will be your trump card?

Ganesh: We have a different approach to revenue generation. On the demand side, we have three products. We have the unified API, Increment, and GoldRush. The way we designed these products is to think of them as multiple ingredients, or the same set of ingredients, but making multiple recipes. Having this structured data, and then having the unified API, Gold Rush which is the blockchain explorer, and Increment, a dashboard product similar to Dune, all based on the same data. That's our approach, getting multiple use cases and personas based on the same indexed data. We have very ambitious goals for revenue generation. We have been growing sequentially for months. As for unlocking, there are some unique opportunities coming up. The first is the RPC-related thing. The current situation is that all the RPC providers are not storing historical archival data. So now the entire industry is consolidating around Covalent to some extent, because the purpose and long-term goal of Covalent is just like EWM, which is to preserve the complete history of the entire blockchain. We have a very customized architecture to achieve this. We have an agreement with Infura. Infura is now starting to direct traffic to us. We can see a few other similar opportunities, I can't name them, but they are all starting to migrate their backends to Covalent as the stack. So we should see significant revenue growth from this move. Beyond that, we have some missing pieces. We've laid this out in our Covalent vision. We've been very candid about some of the gaps that we have in our data stack. One of the biggest gaps is data tracking. Data tracking is a gap that we have for the toughest forensics and accounting cases. We're working hard to make progress and fill that gap. The entire team is structured in such a way that whatever actions they take will drive future revenue growth. This is a completely different part of Covalent, and they are motivated and driven by different reasons. So I'm very confident that we will achieve all of our goals.It's just a matter of product delivery, product pipeline, go-to-market and sales, which again are different parts. I don't think there are many companies in the industry that have this systemic ability to build products outside of the token space.

Thoughts on centralized data indexing

Blair: Thank you for sharing all these insights and behind the scenes stories, it sounds like your mechanism is very sophisticated and well-designed in every way. Looking forward to seeing more innovations from Covalent. Let's look at the bigger picture of data indexing. How do you evaluate the current on-chain data market, focusing on decentralized data indexing? Because there are also centralized data indexes in the market.

Ganesh: Frankly, I think centralized indexers will eventually go away. We saw dozens of indexers come into the market in the last cycle, and most of them are gone now. We are seeing a lot of indexers coming into the market now, and I don’t know what will happen to them. I think centralized indexers are not really in the spirit of decentralized technology, especially if you want to feed this indexed data back into smart contracts. The trust assumption on the data needs to be at the same level as how the data first entered the blockchain. If that trust assumption is broken, then the overall potential size of the market is limited. That’s the thing. For example, if you look at Celestia, Eigen DA, or Avail, the trust in Celestia needs to be exactly the same as the L1 network that is securing it. Otherwise, people will hack Celestia and commit fraud on L1. So it’s important that the setup is the same here. I think centralized indexing service providers may attract some customers, maybe a few million dollars, and may be suitable for some simple use cases. But for the toughest use cases, which is what cryptocurrencies are for, you need trust and security. We never viewed centralized indexing service providers as competition because we have been in this industry for a while and have seen these people come and go and make a lot of noise. Last year Paradigm invested in a project called NXYZ. They raised $40 million, but a year later they collapsed. This happens all the time. We've seen these centralized indexers just don't work in this space. There's a lot of criticism about decentralization when it comes to decentralized indexing service providers. But if you look under the hood, there are some centralized parts, including Covalent. We've been very transparent about our efforts to gradually decentralize. If you're thinking about the vision of who's trying to introduce something completely new, Covalent is the only indexer that has cryptographic security. Every change to the data is cryptographically proven and anyone can audit it. This has been running at scale for years. The first version on the network is launched in the summer of 2022.That was in April, exactly 2 years ago. Despite the normad cross-chain bridge attack and various changes, the network itself has never stopped. So, I think a lot of projects die from lack of focus, not because their core functionality doesn't work. We don’t think about what other projects are doing. We focus on the needs of the industry, the needs of our customers, and the most difficult problems that need to be solved, which will lead the development of the industry. Everything revolves around cryptographic security. When we built all this two years ago, no one asked for this. Two years ago is when we released all this. We've been working on this for four or five years. But that’s what the industry needs, and that’s what’s going to drive this field forward. We are leaders on this journey and it is important for us to let everyone know that.

Blair: I really admire your mentality. I feel the same way because those centralized players sometimes make a lot of noise, but one day it will backfire. I know you don't want me to ask about anything related to investment time, but are there any notable trends that you are watching that you want to share? There is a lot of speculation about this cycle regarding your business data metrics, whether it is the data index volume from Layer1 or any other type of metrics.

Ganesh: I would say Covalent's gains are real. Real customers pay to use the protocol and data. This demonstrates the quality of the data and the quality of the service. No other index currently has revenue of this magnitude. We have clients like Fidelity and Ernst & Young. This tells you how trustworthy the data is. The second one in terms of impact, we did a calculation a few weeks ago and there are over 250 million wallets that use or benefit from Covalent's data. This includes all wallets, all custodians. If you look at Jump's hosting products, they all use Covalent. For example, Ambient Finance is a project on Scroll, AirSwap, and SushiSwap. All these projects use Covalent data to enrich structured data. There are over 250 million wallets in this industry. this is true. This is how many unique wallets we see using Covalent data. Proofs submitted on-chain can be downloaded by anyone and reconstruct the entire Ethereum state from the genesis block. It's real and you don't even have to talk to us. You can just download these proofs and rebuild the entire stack. I would say these are things that really exist on Covalent. In terms of trends, I participated in several panels at the ETH Denver event. I'm pretty sure the cycle is the DA cycle, the data availability cycle. I would say there's also a big push around AI. This is a macro trend and very exciting. I think LSD and LRT seem to be the focus of many people's attention. I'm not a finance person so I don't understand the intricacies of the workings and risk factors of these liquid staking tokens, but it seems like these tokens are getting a lot of attention. Maybe with the Eigen layer and re-staking, there will be big changes this year. But we just focus on the sweet spot where we excel, which is data, data availability and AI.

Blair: Thank you very much for sharing your expertise today. The Web3 industry is still in its infancy. In this field, various changes and adjustments are very frequent. We have seen a lot of fresh capital or innovation pouring into the world, bringing various experiments. Let us wait and see how things develop.

Two suggestions

Ganesh: I want to leave the audience with two thoughts. The first one is, Covalent has about 60,000 developers. Infura probably has about 500,000 developers. So Covalent has about 10% of the people that Infura has. GitHub has 30 million developers, and Covalent has 0.1% of GitHub. We are still in such an early stage, everything is just beginning. The second point is, in the bull market, people will think, should I invest in meme coins? Should I invest in LRT, LST? I would say that the biggest investment you can make is to invest in yourself, invest in your knowledge, invest in your research, and invest in your beliefs. If you believe in yourself, then you should back yourself more firmly. In my limited experience, those who have done this have done well for themselves. So I want to convey this message to our community and our audience.

Blair: That's very true. Thank you so much for everything you shared, very fulfilling.

Ganesh: Thank you so much for the great work you do, and I think we need more of those true builders and people who have strong convictions. If the listeners haven’t had a chance to read it, please read the English and Chinese translation of that study, it’s very extensive, detailed, and goes very deep. Thank you so much for all you do for the industry.