Nansen CEO Talks AI Integration & Future of Blockchain Analytics

Airdrop Is Live CaryptosHeadlines Media Has Launched Its Native Token CHT. Airdrop Is Live For Everyone, Claim Instant 5000 CHT Tokens Worth Of $50 USDT. Join the Airdrop at the official website, CryptosHeadlinesToken.com In an exclusive interview with BeInCrypto, Nansen CEO Alex Svanevik discusses the company’s bold evolution from a traditional blockchain analytics platform to a fully AI-native, multi-chain intelligence hub. With the recent launch of Nansen 2, Svanevik highlights dramatic upgrades in speed, usability, and AI integration that reshape how investors, analysts, and institutions interact with on-chain data.Svanevik also explores Nansen’s strategic expansion, both in terms of blockchain support and geographical presence, while highlighting the platform’s growing validator network, elite research unit, and ambitious plans for embedding AI across the crypto landscape.Svanevik on How Nansen Evolved to Meet User DemandsWe recently completed a major migration from Nansen 1 to Nansen 2, and there were several key improvements we wanted to make. First, we focused on improving latency. The original version had noticeable delays and slow load times, whereas Nansen 2 is significantly faster and more responsive.Second, we received feedback that Nansen could be a bit difficult to use. So with Nansen 2, we made the interface much more streamlined and intuitive.Third, we integrated more AI throughout the product. Users now have several AI-powered features available and AI-driven workflows running in the background.In fact, the entire tech stack was rebuilt from the ground up in Nansen 2.AI Features: Personalized Signals and SearchSignals is essentially a feature that notifies you whenever something unusual happens on-chain. For example, if there’s a token that people are suddenly buying at 100x the usual volume, you’ll receive a signal about it in your feed.What makes Signals powerful is that it’s based not only on what we identify as important and unusual activity, but also on what’s personally relevant to you. We tailor signals based on your interests—looking at your watchlisted tokens, your portfolio, and other preferences—to provide personalized insights.We’ve also made significant improvements to search functionality. You can now simply hit Command + K, type anything you’re looking for, and quickly get relevant results. Looking ahead, we have some new AI features rolling out in the coming weeks and months. The first is an API, which will be extremely useful for developers and advanced users. The second is something we call MCP—Model Context Protocol—which will enable you to use Nansen’s data inside other AI tools like ChatGPT, Cursor, and Claude.This will be a real game-changer. As a journalist or analyst, you’ll be able to just type a question, and the AI agent will fetch real-time, historical on-chain data across any blockchain using our database of nearly half a billion labeled addresses. That gives us the richest on-chain dataset in the world, and puts that power directly at your fingertips.Expanding Beyond EthereumWhen we first started building Nansen, we only supported Ethereum. At the time, I believed that Ethereum would be the chain—possibly the only one people would ever use or that would grow significantly.But today, we live in a highly competitive, multi-chain world. Users now expect to see support for Solana, Layer 2s, Bitcoin, and other blockchains. Because of that shift, we’ve made significant internal investments in our integration processes to dramatically speed things up.What used to take three to six months to integrate a new chain into the product can now, in some cases, be done in just one day. For instance, we were able to integrate BitLayer—a Layer 2 on Bitcoin—relatively quickly. It’s a chain our users have been requesting, and strategically, it aligns well with our long-term vision.Nansen’s Role as a Blockchain ValidatorIt’s not immediately obvious how validators and analytics are connected. But when you look at it from the perspective of Nansen’s user base, the connection becomes clearer.Our users generally fall into two main groups. The first group consists of investors—people who use our platform to gain insight into on-chain activity. The second group includes the blockchain networks themselves—the chains that want to understand what’s happening on their own platforms. To support that, they need to be integrated into Nansen.Running validators allows us to bridge these two customer groups. We can operate validators for the chains we integrate, and at the same time, offer a staking option for our end users.Since acquiring StakeWithUs in September, we’ve scaled our staking business from $60 million in assets staked to over $1.2 billion in just seven months—a 20x increase.So even though, on the surface, it might seem unusual to combine a validator business with an analytics platform, the results speak for themselves. Users are happy to stake with Nansen because they trust the brand and they trust the product.Inside Nansen’s Research PortalWe have a truly elite research team at Nansen, and they operate using what we believe to be the best on-chain data in the world. Ensuring the quality of that data is, of course, absolutely essential.We don’t focus on producing a high volume of research—we prioritize quality over quantity. When you look at some of our standout work, such as our analyses of the FTX collapse or Luna, these are examples of the deep, thoughtful approach we take.In fact, just today I was listening to a New York Times podcast in which the journalists had used Nansen’s data, and our research team advised and supported them during their investigation and production process.It all begins with taking the time to do proper research, backed by great data and fantastic people. On top of that, Nansen is an AI-native organization. Everyone on the team uses AI extensively in their workflows.This has made research even more productive. What used to take hours now takes far less time—with AI tools, a single hour of research can go significantly further than it could two or three years ago.Role of AI in Revolutionizing Blockchain AnalyticsCynics might say that AI and blockchain are just two overhyped technologies that people lump together to raise more money—and to some extent, there’s a kernel of truth in that.However, I tend to think of AI and blockchain as complementary forces.AI creates abundance. It allows for the creation of an infinite amount of content at very low cost. In contrast, blockchains preserve scarcity. They ensure limitations, like the fixed supply of 21 million Bitcoins, for example.So, in many ways, these technologies are opposites that can work in harmony. And there are already concrete examples of how this interplay can solve real problems.We now live in a world where it’s incredibly easy to generate fake content. If you see a photograph, how do you know AI didn’t create it? This is where blockchain can help. You could use cryptographic hashes to verify the authenticity of digital media—to prove, for instance, that a photo is genuine and not AI-generated.​​Identity is another critical area. With so many AI agents now interacting online, it’s getting harder to tell who is human and who is not. That’s why cryptographic identity verification, like what Worldcoin is attempting, becomes increasingly relevant.There’s also a structural contrast: AI has a natural tendency to centralize—large tech companies with vast GPU resources dominate. Blockchain, on the other hand, is fundamentally about decentralization.So, crypto could play a vital role in preserving decentralization, not only in how models are trained but also in how data is collected, verified, and shared. AI in Blockchain SecurityAI plays a major role in detecting scams and identifying problems on the blockchain, and we rely on it heavily at Nansen. In fact, one of our primary uses of AI is address labeling.It’s significantly faster than manual labeling. To put it into perspective, we’ve labeled over 500 million addresses—there’s simply no way humans alone could do that at scale. You’d need a workforce the size of an entire country to achieve that manually.Instead, we use finely tuned AI agents to perform the labeling work. At the same time, we have rigorous quality assurance processes in place to ensure the AI isn’t hallucinating or assigning incorrect labels. This balance allows us to scale responsibly and maintain high accuracy.User Privacy vs. Transparency At NansenI think this presents a very interesting ethical dilemma. The reality is that blockchains are inherently transparent. At Nansen, we see our role as somewhat journalistic—these events and data points are happening on-chain, and our responsibility is to organize and surface that information in a meaningful way.The data we use to label addresses comes entirely from the public domain. We don’t have privileged access to any private or off-chain information. Instead, we rely on a combination of on-chain patterns and publicly available sources on the web. Our systems analyze these inputs and perform inference-based labeling to identify the likely owners of addresses.Of course, we are fully compliant with global privacy regulations. For example, if an individual publicly discloses their wallet address—perhaps in an online forum post about participating in governance—and our AI picks it up, they can later request that we remove the label. In such cases, we honor those requests.However, the same rules don’t apply to corporations or investment funds. As non-individual entities, they do not have the same privacy rights, and we are not obligated to remove those labels. Helping Regulators and Companies Leverage Blockchain DataI’d say our core focus has always been on crypto-native investors and funds—that’s the audience we’ve built Nansen for, and continue to prioritize.We haven’t invested much time engaging directly with regulators or government entities. That said, I’m aware that some regulators do use our products. We don’t actively market or sell to them, but of course, they can subscribe like any other customer.In fact, one of the top exchanges told me that VARA had once contacted them about something they had seen on our platform. We’ve never had a sales meeting with VARA, so it’s clear that they’re using Nansen organically—and that’s quite interesting.I do believe there’s a lot more regulators could do by leveraging on-chain data.For example, in 2022 during the collapse of Terra, there was a Singapore-based company called Hodlnaut, which operated similarly to BlockFi. We were able to see on-chain exactly what they were doing with customer funds—they were allocating almost all of it to Anchor, the Terra DeFi protocol that promised 20% yields.Meanwhile, they were only offering their users 8%, keeping the difference as revenue. The problem arose when UST de-pegged—resulting in hundreds of millions of dollars in losses. We saw it happening in real time.If regulators had monitored on-chain activity, they might have caught it sooner. I understand that many don’t, because blockchain analysis is still relatively new and technically complex. So I don’t blame them, but I do see a clear opportunity for better real-time oversight.I’ve actually recommended to regulators in Singapore that any licensed exchange should be required to disclose all of their wallet addresses. That way, regulators can monitor all on-chain flows—both into and out of the exchange.Everything is transparent. They don’t need to use Nansen specifically—they could use any blockchain analytics platform. But in my view, disclosing wallet addresses should be part of the licensing process. It gives regulators far more visibility into how exchanges operate.And another case from 2022 is also worth mentioning—the FTX collapse.Even though Sam Bankman-Fried publicly claimed that withdrawals were frozen, we saw funds moving out of FTX wallets. When we asked about it, we discovered that withdrawals had been reopened—but only for Bahamas-based KYC users. He reportedly did this out of concern that local authorities might arrest him.So people began buying Bahamas KYC accounts on the black market to get their money out.These are exactly the kinds of events that a sophisticated regulator—like VARA, for example—could monitor much more effectively using on-chain data.The thing I’m most excited about right now is our upcoming AI features.As I mentioned earlier, we’re releasing a new API in beta in the coming weeks, and alongside that, we’re launching something called a Model Context Protocol, or MCP. For those unfamiliar, MCP is essentially a standardized interface for AI tools and products.What that means is you’ll be able to integrate Nansen into existing AI environments like ChatGPT, Cursor, or Claude—and interact with it as if you were speaking to a research analyst or an entire research team. The agent will have access to all of our data, and you won’t need to manually navigate the platform to find the answers you need—the AI will handle that for you.A recent example really illustrates the potential. I saw a fund arguing with a token project on Twitter, and I wanted to understand the background. I simply asked our AI assistant to provide the full trading history of that fund with that token. Within 30 seconds, it generated a one-page summary of all relevant transactions and context. That kind of research would have taken me a long time to do manually.So to me, the MCP is incredibly exciting—it’s going to change how users interact with blockchain data.And looking ahead, in Q3, we’ll be launching a completely redesigned Nansen interface. It will be agent-first and AI-native, moving away from the traditional dashboard-and-click model. We’re aiming to deliver a much more intuitive, intelligent user experience—one that feels like having a personal on-chain assistant at your fingertips.Nansen’s Unique EdgeThere are a few key things that set us apart.First, we have what we believe is the best on-chain data in the world. We support a broad range of blockchains, and more importantly, we offer the highest-quality and most comprehensive labeling. That means we can accurately identify which addresses belong to which entities.But we go even further. We also calculate behavioral metrics—like profit and loss (P&L)—which brings me to the second point: our 100% focus on traders and investors.These users want to know what smart money is doing. So we’ve done the work of identifying and classifying which wallets qualify as smart money, by analyzing their trading history and calculating their P&L, which is no small task. This allows us to say: “Here are the 5,000 smartest wallets in the world—and here’s what they’re buying today.”So between the quality and depth of our data, and the fact that it’s purpose-built for traders and investors, we believe we offer something truly unique.Lastly, I’d add that we’ve designed Nansen to be a full-stack product. You don’t need multiple platforms to manage your on-chain activities. With Nansen, you can track your portfolio, set on-chain alerts, perform due diligence, discover and research tokens, and even stake assets. And in the future, you may even be able to trade directly within Nansen.Ensuring On-Chain Data AccuracyWhen it comes to our labels and attribution data, one of the most important things is that we have evidence for every single label in our database. As I mentioned earlier, we’ve labeled nearly 500 million addresses, and for each one, whether it’s labeled as Binance, Sam Bankman-Fried, or any other entity, we provide verifiable evidence in our system.So if there’s ever a mistake or dispute, you can go into our database and read exactly why a label was applied. For instance, there was one case where an address was labeled as the Luna Foundation, and it later turned out to be Do Kwon’s personal wallet—two things that are obviously very closely related. Was that technically a mistake? Perhaps. However, the supporting evidence clearly explains the rationale behind the label, which helps users understand the context.This level of transparency and traceability is, I believe, quite unique among blockchain analytics platforms.Another thing that sets us apart is that we’ve developed our own AI benchmarking system. Just like how a new LLM (large language model) is evaluated based on how well it performs in math, reasoning, or coding—we have our own internal benchmark for how well an LLM can label blockchain addresses using our system.So, when a new model comes out from OpenAI, Anthropic, or others, we can immediately test it against our ground truth dataset—thousands of pre-verified addresses—using our custom prompts and infrastructure. This tells us whether the new model is more accurate, less accurate, or just different.It’s a very robust quality assurance process, and it’s essential because the hardest part isn’t the raw on-chain data—that’s usually easy to verify. You can cross-check it with multiple sources.But the real challenge is in accurate labeling and attribution—and that’s where our evidence-backed system and AI benchmarking give us a clear edge.Global Expansion PlansThe US has always been Nansen’s largest market since the beginning. That said, I believe the US is becoming even more pro-crypto, and we’re aligning with that momentum.We’ve recently opened an office in New York, and we’re also actively hiring there. If I were to choose the next location for expansion, it would most likely be Dubai. There’s a lot of exciting activity happening here, and many strong crypto companies are already established in the region.Of course, there’s significant regulatory support through forward-thinking licensing regimes. But beyond that, there’s also a strong message of openness—Dubai has made it clear that crypto companies are welcome.We don’t require a license ourselves, since we’re not a centralized exchange, so it’s not an immediate regulatory need. But I still think having a local presence in Dubai would be incredibly valuable—it’s simply a great place to be right now for the crypto industry.I recently visited the Bybit office here, which is impressive—it’s a large space, and they even have their own branded car! It really reflects how much the crypto sector is growing in this region.One thing I do wonder about, though, is how fast Dubai is growing overall. Last time I was here, I heard that 1.5 million people moved to Dubai in the past year. I’m not sure if that number is accurate, but it certainly feels like the city is expanding rapidly.What Industry Leaders Should Focus OnMy main piece of advice, based on what I’ve learned over the years, is that the most important element in any company is its culture.It’s critical to clearly define the kind of culture you want to build, and then actively live by it. That means ensuring that everyone in the company understands and embodies those values in their day-to-day work.You should recognize and reward the people who are strong advocates and ambassadors of that culture. And sometimes, you may need to make tough decisions, including parting ways with individuals who aren’t aligned.DisclaimerIn compliance with the Trust Project guidelines, this opinion article presents the author’s perspective and may not necessarily reflect the views of BeInCrypto. BeInCrypto remains committed to transparent reporting and upholding the highest standards of journalism. Readers are advised to verify information independently and consult with a professional before making decisions based on this content. Please note that our Terms and Conditions, Privacy Policy, and Disclaimers have been updated.Source link