Dune has cut 25% of its workforce as the crypto data platform sharpens its focus on artificial intelligence, core blockchain analytics products and institutional demand for onchain data.
Co-founder and CEO Fredrik Haga said the company is restructuring to concentrate on the data products that crypto teams rely on every day. He confirmed that one-quarter of the team was let go this week, while also saying the company remains well capitalized and is moving deeper into two major areas: AI-powered analytics and institutions coming onchain.
The move is not being presented as a retreat from the market. Instead, Dune is framing the restructuring as a strategic reset. The company appears to be betting that the next stage of crypto analytics will not be built only around human analysts writing SQL queries and publishing dashboards. It will be built around AI agents that can search, query, analyze and visualize blockchain data with far less manual work.
That shift could change what users expect from Web3 analytics platforms. In the past, Dune became popular because it made blockchain data searchable and shareable through public dashboards. Now, the company is trying to make that same data usable by AI systems, financial institutions and enterprise teams that need faster answers from increasingly complex onchain markets.
Why Dune’s AI Pivot Matters
Dune has become one of the best-known names in blockchain analytics. Its homepage describes the platform as an onchain data provider trusted by more than 20,000 companies, with support for more than 100 chains, over 1 million users and more than 1.5 million datasets.
That scale explains why the layoffs are being watched closely. Dune is not a small startup quietly reducing headcount after losing product-market fit. It is a widely used crypto data platform making a deliberate decision to operate with a smaller team while investing more heavily in automation, AI agents and institutional services.
At the center of that strategy is Dune MCP, a Model Context Protocol server designed to connect AI agents to Dune’s data warehouse. Dune says the MCP server gives agents structured access to raw and decoded data across more than 100 blockchains, curated datasets and community-built tables. It can discover tables, write DuneSQL queries, execute them, retrieve results and generate visualizations from a single conversational workflow.
In simple terms, Dune wants users to ask questions in natural language and let AI handle the data work in the background. Instead of manually searching for tables, writing SQL, debugging errors and building charts, a user could ask an AI agent to analyze stablecoin flows, compare decentralized exchange volumes, track real-world asset tokens or monitor wallet behavior.
That is a major change for crypto research. The old workflow rewarded analysts who knew where the data lived and how to query it. The new workflow may reward teams that know how to ask better questions, verify AI-generated results and turn analysis into decisions.
AI Agents Are Becoming the New Interface
Dune has already described itself as becoming “agent-native.” In March, the company launched Dune CLI and Skills, saying AI agents are becoming the primary interface for onchain data because they do not click through dashboards; they query, execute and return results programmatically. The company said its CLI gives AI agents direct terminal access to Dune and supports workflows such as dataset discovery, DuneSQL execution, query management, documentation search and usage tracking.
The company’s AI agents page also emphasizes the same direction, saying users can plug Dune into any MCP-compatible AI client or connect through the CLI to give an agent access to live onchain data. It says Dune’s AI tooling is designed for programmatic access, structured machine-readable responses, parameterized queries, async execution and enterprise controls.
For crypto companies, this could be useful. A DeFi protocol might want an AI agent to monitor liquidity pools and flag abnormal withdrawals. A venture fund might want daily summaries of token flows across Ethereum, Solana and Base. A compliance team might want quicker visibility into wallet movements or stablecoin activity. A researcher might want to compare trading volume across dozens of decentralized exchanges without writing every query manually.
This is the promise behind AI crypto analytics: faster workflows, broader data access and fewer technical bottlenecks. But it also creates a harder question for workers. If AI can automate parts of data discovery, dashboard creation and research operations, companies may need fewer people for some tasks while hiring more specialists for others.
The Layoffs Fit a Larger AI Workforce Trend
Dune’s decision comes during a wider wave of job cuts tied to AI investment and operational restructuring. Reuters reported that concerns are growing among investors and economists that AI will disrupt established industries, with job losses already appearing in sectors most exposed to automation. Reuters also cited data from Challenger, Gray & Christmas linking AI to 7% of total planned U.S. layoffs announced in January.
The Associated Press has also reported that more companies are pointing to AI when announcing cuts, even when AI is not the only reason. AP noted that some businesses are reducing staff as they redirect spending toward AI, streamline operations or reallocate resources, while still describing the explanations as often vague.
That distinction matters. It would be too simple to say AI alone caused Dune’s layoffs. The more accurate reading is that AI is changing how the company wants to build and sell its products. Dune is not just cutting costs. It is reorganizing around a belief that blockchain data will increasingly be consumed by agents, institutions and automated workflows rather than only by human dashboard builders.
Institutional Onchain Data Becomes a Bigger Prize
The second part of Dune’s strategy is institutional crypto data. Haga has linked the restructuring to a future where more financial assets move onchain, including currencies, stocks, bonds and commodities. The company plans to invest in its data layer and higher-touch services for financial firms that need reliable blockchain analytics.
This is a logical market to chase. As tokenized assets, stablecoins and blockchain settlement systems grow, financial institutions need cleaner data, better APIs, compliance-ready reporting and support that feels closer to enterprise software than community dashboards.
Dune’s current product lineup already points in that direction. Its homepage highlights APIs, connectors and DataShare tools that let teams bring onchain data into their own environments. It also describes Dune as an enterprise platform for teams that need to analyze, build and scale with blockchain data.
The opportunity is clear: if banks, funds, payment companies and asset managers move more activity onto blockchain rails, onchain data becomes business-critical infrastructure. The challenge is that institutional customers demand reliability, documentation, performance, support and security. Serving them may require a different company structure than serving open crypto communities.
A Hard Reset for a Changing Market
For employees affected by the layoffs, the strategic logic does not soften the immediate impact. A 25% reduction is a major cut, and it shows how quickly AI is reshaping even companies built around technical communities.
For Dune, the bet is that a smaller, more focused team can move faster in a market where AI tools are changing the economics of software work. If the strategy works, Dune could become a central data layer for AI agents, crypto analysts and institutional onchain finance. If it misfires, competitors may use the disruption to win users who still want deeper human support, specialized dashboards or alternative data systems.
The broader message is difficult to ignore. Crypto data is becoming more valuable, but the way that data is produced, queried and delivered is changing. Dashboards are not disappearing, but they may no longer be the main interface. SQL is not becoming irrelevant, but fewer users may need to write it directly. Analysts are not being replaced completely, but their work is shifting toward verification, interpretation and strategy.
Dune’s layoffs mark one of the clearest signs yet that the AI transformation has reached Web3 analytics. The company is betting that the future of blockchain data will be faster, more automated and more institutional. The cost of that bet is a smaller workforce today, and a very different vision of how crypto intelligence will be built tomorrow.