Rosa Del Mar

Daily Brief

Issue 89 2026-03-30

Roadmap And Market Forecasts

Issue 89 Edition 2026-03-30 9 min read
General
Sources: 1 • Confidence: Medium • Updated: 2026-04-11 19:30

Key takeaways

  • Nansen is developing “Nansen 3” aimed at mass-market consumers and claims it will be the best product Nansen has built, with a futuristic but simple user experience.
  • Nansen has a data platform with more than 500 million labeled blockchain addresses and real-time on-chain flow analytics.
  • Nansen decided to vertically integrate toward investors/traders rather than remain a horizontal analytics provider, aiming to become a full-stack on-chain agentic trading product.
  • Nansen believes agent behaviors will diverge across users due to differing intent, risk tolerance, portfolios, data access, tools, model choices, and LLM stochasticity.
  • Svanevik says Nansen is unlikely to launch an AML/compliance product despite having relevant data, because it would distract from serving crypto-native investors and traders.

Sections

Roadmap And Market Forecasts

  • Nansen is developing “Nansen 3” aimed at mass-market consumers and claims it will be the best product Nansen has built, with a futuristic but simple user experience.
  • Svanevik says there is an upswing in open-source AI activity and calls for broader participation in open-source AI experimentation, drawing parallels to early DeFi.
  • Nansen is working on “Smart Money 2.0” intended to predict which addresses are likely to make money in the future, with release expected in weeks rather than months.
  • Nansen plans a more visual agent UX using a curated library of “artifacts” and a “Dunstan 3” release later this year that generates question-specific visualizations in sub-second time.
  • Svanevik predicts trading will be dominated by agents, potentially reaching millions or billions of trading agents within a couple of years and becoming a default mode by around 2028.
  • Nansen expects users to shift from micromanaging individual trades to specifying higher-level strategies for agents to execute.

Attribution As Data Moat And Pipeline

  • Nansen has a data platform with more than 500 million labeled blockchain addresses and real-time on-chain flow analytics.
  • Nansen outsources node provisioning to providers such as QuickNode and Alchemy while indexing data in-house into ClickHouse.
  • Svanevik argues that raw on-chain data is commoditized but the enriched attribution layer is difficult to recreate and drives demand for Nansen labels.
  • Nansen’s on-chain analytics stack includes indexing raw chain data into ClickHouse, harmonizing schemas across chains/DEXs, and enriching data with an attribution layer that labels addresses.
  • Nansen states it labels identities only from public-domain information and will remove labels upon request for individuals (but not for projects/corporations), while noting some on-chain linkages (e.g., ENS) are immutable.
  • Nansen labels deterministic on-chain entities (e.g., Uniswap pools) via on-chain events and infers off-chain-controlled entities (e.g., exchange wallets) using behavioral flow analysis and hybrid heuristics.

Vertical Integration Research To Execution Agent Product

  • Nansen decided to vertically integrate toward investors/traders rather than remain a horizontal analytics provider, aiming to become a full-stack on-chain agentic trading product.
  • Nansen relies on partners for trading infrastructure including Privy for embedded self-custodial wallets and DEX aggregators such as LiFi, OKX DEX API, and Jupiter (Solana) for swap routing.
  • Svanevik claims the Nansen agent can answer a wide range of on-chain questions using near-real-time data (seconds-old).
  • Nansen describes a “trust ladder” for agentic trading that starts with user-confirmed trading and progresses toward auto-execution and strategy-level delegation once reliability is proven.
  • In Nansen’s current mobile app flow, the user must explicitly execute each trade via a confirmation modal and a biometric MFA step.
  • Nansen’s mobile app (Nansen AI) allows users to chat with an agent for on-chain analysis and execute token purchases in the same product.

Agent Training Validation And Personalization

  • Nansen believes agent behaviors will diverge across users due to differing intent, risk tolerance, portfolios, data access, tools, model choices, and LLM stochasticity.
  • Nansen is pursuing “co-creation” where users describe a strategy and the agent provides performance feedback by testing whether it would have worked over past periods.
  • Nansen Gym trains trading agents by replaying historical on-chain data as a fast-forwarded “time travel” environment where the agent is unaware it is seeing a replay.
  • Svanevik claims a vanilla LLM is insufficient for on-chain use because it lacks real-time blockchain data, cannot reliably execute/sign transactions without tool integrations, and often misjudges significance without domain tuning.

Boundaries And Normative Positions

  • Svanevik says Nansen is unlikely to launch an AML/compliance product despite having relevant data, because it would distract from serving crypto-native investors and traders.
  • Svanevik argues that using AI as justification for workforce cuts is often a sign a company was poorly run or lacked the right people, because AI can increase the economic value of productive teams and make hiring more rational.
  • Svanevik argues that AI moratoriums or heavy overregulation in the West could be counterproductive because others will build AI anyway, implying an arms-race dynamic.
  • Svanevik believes the risk of AI causing total human wipeout is real and suggests a best-case long-run scenario may involve humans merging with machines.

Watchlist

  • Svanevik says there is an upswing in open-source AI activity and calls for broader participation in open-source AI experimentation, drawing parallels to early DeFi.
  • Nansen is developing “Nansen 3” aimed at mass-market consumers and claims it will be the best product Nansen has built, with a futuristic but simple user experience.

Unknowns

  • What is the independently verified accuracy of Nansen’s labels (e.g., precision/recall) across major entity categories such as exchanges, funds, and market makers?
  • How frequently are labels corrected or removed, and what is the operational process and turnaround time for disputes/takedown requests?
  • What are the unit economics of the new monetization model (subscription revenue vs trading fees vs staking revenue), and what share of users convert into meaningful trading volume?
  • What is the measured execution quality of Nansen-routed swaps (slippage, price improvement, failure rates) relative to common benchmarks, and how does routing differ across aggregators?
  • What are the reliability metrics and failure modes for Nansen’s near-real-time analytics (latency, chain reorg handling, data completeness), especially given outsourced node provisioning?

Investor overlay

Read-throughs

  • If Nansen 3 ships with simpler UX, Nansen could expand from crypto native analysts into mass market retail, potentially increasing subscriptions and widening the funnel into execution products.
  • Vertical integration from analytics to agentic trade execution could shift monetization mix toward trading linked revenue, making growth more sensitive to execution quality and partner routing reliability.
  • A large labeled address dataset plus near real time flow analytics may support defensible differentiation, but only if label quality and data reliability are measurably high at scale.

What would confirm

  • Public release of Nansen 3 on schedule with observable adoption, retention, and conversion from analytics usage into execution usage.
  • Disclosed or independently assessed execution quality metrics for Nansen routed swaps such as slippage, failure rates, and price improvement versus common benchmarks.
  • Transparent label QA metrics and dispute processes, including precision and recall by entity category, correction frequency, and turnaround time for takedowns.

What would kill

  • Nansen 3 delays or limited uptake, indicating mass market positioning does not translate into sustained engagement.
  • Consistently worse swap outcomes or elevated failure rates versus benchmarks, or partner dependencies that create reliability issues for execution flows.
  • Evidence of low label accuracy, slow or inconsistent correction workflows, or unreliable near real time analytics such as latency, reorg issues, or incomplete data.

Sources

  1. 2026-03-30 podcasters.spotify.com