Alex Svanevik, CEO and co-founder of Nansen
Singapore-based Nansen has spent years helping investors make sense of on-chain activity, turning raw blockchain data into signals about capital flows, behaviour, and market structure. The GIC- and Andreessen Horowitz-backed firm has just launched an AI trading agent designed to operate as an agentic system: ingesting real-time on-chain data, reasoning over live market conditions, and executing trades directly rather than stopping at insight generation. With this move, Nansen is pushing beyond analytics into execution, introducing an AI-driven system that can interpret live on-chain signals and act on them directly.
e27 spoke with Alex Svanevik, CEO and co-founder of Nansen, about what this shift says about the maturity of crypto markets, the evolving role of AI in trading, and how access to institutional-grade workflows could reshape who gets to move first on-chain.
Excerpts from the interaction:
You’ve positioned Nansen as moving from “pure analytics” to “agentic trading.” What was the key inflexion point that convinced you the market was ready for this shift? How do you see this repositioning affecting your competitive moat against other on-chain analytics platforms?
The inflexion point was realising that insights alone weren’t enough anymore. Users could see opportunities on Nansen, but they still had to leave our platform to act on them. At the same time, AI matured to the point where it could safely reason over live on-chain data in real time.
Also Read: ‘The future is on-chain’: Nansen CEO on AI, staking, and new growth plans
By closing the loop between insight and execution, we’re not just adding a feature; we’re changing the workflow. Our moat deepens because this is built on Nansen’s proprietary 500 million-plus labelled wallet dataset, which competitors can’t easily replicate.
A press release notes that Singapore, a major crypto hub, is restricted from accessing Nansen’s trading features. How does this regulatory limitation impact your growth strategy in Southeast Asia, and what’s your timeline for expanding into these restricted markets?
Our user base is global by nature, with the US as our largest market, followed by established markets such as Japan, South Korea, and Hong Kong, and strong momentum in emerging markets, including Vietnam, Indonesia, and India.
Nansen is fundamentally an on-chain, non-custodial product. It puts us in a different category from traditional custodial exchanges or brokers, as we’re fundamentally a technology provider.
Furthermore, regulatory interpretations can vary significantly across jurisdictions. In some markets, including Singapore, we’ve taken a deliberately conservative approach by limiting certain functionality while we focus on our core strengths: on-chain analytics and portfolio tracking.
From a growth perspective, this hasn’t been a constraint. Southeast Asia is an important region for us, and our growth there today stems from our leadership in on-chain analytics. When it comes to expanding functionality in more tightly regulated jurisdictions, we take a long-term view.
How does the introduction of trading functionality change Nansen’s monetisation strategy? Will you maintain a freemium model, introduce trading fees, or adopt a hybrid approach?
Our goal has always been to surface the signal and create winners with our product.
Today, we’re making our Solana analytics available to our free users as well. This is part of a move to make Nansen accessible to more people.
Furthermore, we recently simplified our pricing to a single plan, Pro, at US$49/month. Previously, our users were paying up to US$999/month for the same features, so they now literally get 20x the value. We are playing the long game and wanted to make Nansen accessible to more people.
Also Read: On-chain analytics firm Nansen acquires staking provider stakeWithUs
The Pro plan gives you the best on-chain analytics, the best on-chain agent, and the lowest trading fees at 0.1 per cent.
With over 500 million labelled addresses as your competitive advantage, how do you prevent competitors from building similar datasets, and how frequently do you need to update and validate this data to maintain accuracy?
This dataset is the result of years of continuous labelling, behavioural analysis, and validation across multiple market cycles. We update and re-validate data continuously because on-chain behaviour changes every day. The real advantage isn’t just scale; it’s accuracy, context, and history.
How do you differentiate Nansen AI from established crypto trading platforms like Kraken, Coinbase, or Bybit that are also integrating AI features? What’s your unfair advantage?
Those platforms start with custody and execution and then add analytics. We start with intelligence. Nansen AI is built on live, on-chain behavioural data that shows who is moving capital, where, and why. That allows us to reason about markets in a way that price-centric platforms can’t. Trading is a natural extension of that intelligence, not an add-on.
You claim Nansen AI outperforms ChatGPT, Claude, and Grok on on-chain benchmarks. What specific benchmarks are you using? Rather than positioning Nansen AI against general-purpose models, could you integrate with ChatGPT or Claude plugins to reach a broader audience? What are the trade-offs you’ve considered?
We tested this in real-world on-chain questions: wallet identification, token discovery, due diligence, and portfolio analysis. Nansen AI hit over 85 per cent quality scores in Expert Mode. GPT-5? Around 20 per cent.
Standard LLMs don’t understand wallet labelling or trading patterns. We actually use foundation models as part of our stack; we constantly benchmark them, then fine-tune and extend them with on-chain capabilities. We even route certain prompts to specific models based on what they’re optimised for. But the intelligence is baked into our system, not just retrieved from a data layer.
The press release emphasises that “the agent operates under user-defined rules” and “does not take custody of funds.” How do you balance giving users powerful, autonomous trading capabilities with this commitment to transparency and control?
The agent is powerful, but the user is always in control. Every action is based on user-defined rules, fully transparent, and executed on verifiable blockchain data. We don’t take custody of funds, and users can intervene or stop the agent at any time.
What’s your user acquisition strategy for this new trading product? Are you targeting retail traders migrating from Coinbase/Kraken, or institutional investors looking for sophisticated analysis tools?
Users already come to Nansen to discover and research what’s happening on-chain. We are already a destination. Now, we’re making the user experience more streamlined and integrated so you can execute your trades via our interface. By partnering with the best DEX aggregators in the world, we offer optimised routing and fees as low as 0.1 per cent. Why would users trade anywhere else?
Beyond trading, what other on-chain actions do you envision your AI agent performing in the next 12-24 months? Are you exploring lending, staking, governance participation, or other DeFi primitives?
Our product vision is simple: trade everything on-chain, with AI. This is what we call A New Way To Trade. We’ve thought deeply about what trading looks like in 2030. Instead of waiting until then, we decided to build that user experience today.
Also Read: Nansen raises US$75M to help users make informed decisions before investing in crypto, tokens
In the near term, we will bring our AI agent to the web; currently it’s only accessible via our mobile app.
Nansen also offers staking, which grew from US$60 million to US$2 billion AUM in just nine months. We will continue to expand the chains we support around this, allowing users to do everything on-chain from Nansen: discover, decide, trade, stake, and track.
The post Why Nansen believes AI agents are the future of on-chain markets appeared first on e27.