Dune, the crypto data platform, has reduced its workforce by 25% as it concentrates on artificial intelligence tooling and servicing institutions that are moving onchain. CEO and co-founder Fredrik Haga announced the decision on Thursday, May 14, 2026, framing the cuts as a way to narrow attention to the company’s core data products and accelerate work on Dune MCP—its new tool for assembling analytics dashboards without requiring SQL skills or specialized data infrastructure expertise.

Technology Overview

Dune positions itself as an end-to-end provider of blockchain data services for thousands of users across the industry. In announcing the restructuring, Haga characterized the company’s value proposition as having built the full stack for crypto data—from raw chain information to tools that enable practical analysis and visualization. That framing underscores Dune’s effort to deliver a consistent pathway from onchain activity to decision-ready insights, particularly for teams that may not have dedicated data engineering resources.

A centerpiece of that approach is Dune MCP. As described by the company, MCP is intended to let users build dashboards without knowledge of SQL or the mechanics of data pipelines. For Web3 teams, that simplifies routine analytics tasks: querying common metrics, assembling protocol- or token-level views, and sharing results across an organization. By lowering the technical bar for analysis, the product is designed to shift more of the work to interfaces and templates rather than database syntax, which aligns with the platform’s push to broaden access while keeping analysis grounded in onchain data.

How It Works

While Dune has not detailed the internal mechanics of MCP in this announcement, the concept suggests a model where complex steps—such as extracting onchain events, normalizing schemas across networks, and indexing contract interactions—are abstracted behind a guided workflow. In practice, this kind of system typically presents predefined queries, parameter inputs, and visualization modules that convert raw ledger activity into charts and tables that non-specialists can interpret. That design aims to make routine analytics faster to produce and easier to maintain as protocols evolve.

Haga said Dune is “all-in on two shifts: AI and institutions coming onchain,” describing a strategy that uses AI to streamline how people create and refine analytics while addressing needs specific to larger, compliance-minded organizations. In a data platform context, AI can assist with query generation, recommend relevant datasets, and help validate outputs against known patterns. For institutions, the emphasis often falls on standardized views, reliability, and operational rigor—areas where an integrated stack can reduce friction compared with assembling tools from multiple vendors.

Industry Impact

The changes at Dune come amid a broader recalibration across crypto companies toward leaner operations and AI-enabled workflows. Over recent months, several well-known organizations have also announced staff reductions, citing a combination of efficiency drives and evolving operating models. Coinbase said last week that it would cut 14% of its workforce, with CEO Brian Armstrong describing the move as part of an AI-first structural overhaul. The Algorand Foundation reduced headcount by 25% in March, and Gemini trimmed a similar share in February while exiting the EU, UK, and Australian markets.

Hiring patterns are shifting alongside these reductions. In March, Haga said Dune conducted 300 interviews for engineering and data roles and set “AI fluency” as a non-negotiable criterion. The message was unambiguous: candidates were expected to be hands-on with AI and actively exploring how it could reshape their domain. That stance reflects a wider sentiment that teams should be smaller but more capable, with AI multiplying the output of fewer people rather than spreading work across larger staffs.

Recruiters and founders observing the market have described a simultaneous push for automation and discipline. Robert Lycett, Head of Recruitment at RiskPod, said AI is clearly reducing costs at the junior level, noting that low-level administrative work can be automated in ways that deliver immediate payroll savings. At the same time, he cautioned that AI can serve as a convenient justification for cuts that are also driven by market conditions, with companies prioritizing profitability and efficiency over growth in a period where expansion has slowed.

Other voices emphasize that jobs are not disappearing wholesale but being re-scoped. Vedang Vatsa, founder of HashtagWeb3.com, said teams are trimming basic technical roles to stay lean while paying a premium for engineers who can integrate AI into blockchain products. That pattern tracks with what many Web3 projects require: the ability to bind user-facing features to onchain events, all while embedding AI-driven assistance or automation into the data flow.

Shubhada Pande, founder of Art of Blockchain, framed the trend as a redesign of how work gets done rather than a one-to-one replacement of Web3 jobs by AI. In her view, the roles most vulnerable are those tied to repetitive or low-impact tasks, while core positions are evolving to include working alongside AI tools, validating machine-generated output, and applying human judgment where automation falls short. The result, she said, is a higher bar for what constitutes a sustainable role in the sector.

Future Implications

Dune’s strategy signals where a significant slice of Web3 infrastructure may be heading. If MCP and similar tooling can reliably convert onchain data into accessible dashboards, teams across exchanges, protocols, and analytics firms could spend less time on query syntax and middleware and more time on interpreting signals and shipping product. Combined with an AI-first stance, that workflow promises shorter cycles for analysis, documentation, and reporting—especially as organizations seek institutional-grade operations.

For institutions coming onchain, the immediate priorities often include clarity, repeatability, and auditability of analytics. A platform that assembles these capabilities within one stack may reduce the need for bespoke integrations and manual data stitching. As firms lean into AI for drafting, classification, or summarization tasks, the importance of robust base data increases: automation amplifies whatever foundation it is given. In that context, Dune’s focus on core data products mirrors a wider industry movement to standardize and harden the layers that power dashboards, monitoring, and decision-making.

Despite the staff reductions, Haga said the company remains well capitalized and focused on long-term development. He pointed to Dune’s ability to navigate multiple industry cycles over the past eight years and reiterated the company’s conviction in maintaining the flow of data to customers across the crypto ecosystem. The current reset suggests a near-term emphasis on disciplined execution: fewer people, sharper priorities, and tools designed to help both Web3 teams and institutions translate blockchain activity into actionable insight.

Across the sector, the message is consistent. Companies are seeking to do more with less, make analytics easier to access, and apply AI where it can shoulder routine work while leaving specialized judgment to human experts. For a data platform like Dune, consolidating around an end-to-end stack and lowering the entry barrier for analysis aligns with those aims. As organizations deepen their onchain operations, the demand for dependable, intelligible, and institution-ready analytics is likely to define the next phase of crypto data infrastructure.