Background & Biography
Mats Olsen co-founded Dune with Fredrik Haga in 2018, after working at Schibsted. Dune's own history places the company's beginning in Oslo, and Norwegian reporting independently identifies both founders and their earlier employer. In an AWS article co-authored by Olsen, the founding problem is described in engineering terms: building with Ethereum exposed how difficult it was to obtain structured information from blockchain records. Public availability alone did not make those records convenient to analyze.
Dune developed a platform through which users could query data and share the results as dashboards. Its early application was rebuilt as the company added staff in 2021, according to Haga's detailed account of that period. The company later raised a $69.42 million Series B, announced in February 2022, giving the team more resources to develop its tools. These were collective company milestones, not evidence that Olsen personally implemented every feature or that financing itself proved the technology's quality.
Olsen's current CTO title is explicitly confirmed in his May 2026 announcement about Sim's closure. His public professional profile also describes him as Dune's co-founder and CTO and identifies Zug, Switzerland. Older material placing him in Norway belongs to an earlier period; it should not silently override more recent professional information. The useful geographic context is his Swiss professional base and the company's Norwegian origins, without making claims about citizenship or private residence.
Notable Contributions
One of Olsen's clearest documented contributions is explaining and leading Dune's technical product direction. His March 2023 introduction to DuneSQL described a transition away from earlier PostgreSQL and Spark SQL experiences toward a query engine built on Trino. The announcement covered migration tooling, changes in SQL behavior and data types suited to blockchain analysis. It also acknowledged that communication about the transition had been insufficient, illustrating that an infrastructure migration is a user-coordination problem as well as a software change.
Current Dune documentation confirms the continuing place of DuneSQL and its blockchain-specific functions. Binary data, address encodings and large integer values are not peripheral concerns when analyzing smart-contract records. Supporting them in the query layer makes it easier to express the question an analyst actually wants to ask. Olsen's public technical leadership is therefore relevant beyond speed: the engine's abstractions influence whether analysts can represent transactions accurately and move existing work between versions.
The data-processing layer is only part of the system. Spellbook, Dune's open repository of analytical models, transforms raw and decoded records into reusable datasets through a dbt project. Its current repository documents testing, dependencies and contribution workflows. Those models are work by a distributed set of contributors, and should not be reassigned to the CTO merely because he leads the company technically. His role is better understood in relation to the platform that supports collaboration and makes those contributions usable together.
A 2023 AWS account gives another view of his responsibilities. Olsen described consolidating cloud infrastructure and using services for storage, processing and application delivery, including Amazon S3 and Kubernetes-based infrastructure. He also discussed compute costs and the importance of matching infrastructure spending to customer value. The article is a supplier case study, not an independent performance audit, but it offers a specific record of technical choices and the operational problems they were intended to address.
Dune's API extended the platform beyond its website. A December 2022 product presentation featured Olsen introducing the ability to turn queries into endpoints for other applications and workflows. That direction matters because useful analysis need not end at a chart: another system may need to retrieve the same results repeatedly. An API also changes the obligations of a data provider, making refresh behavior, reliability and compatibility important to users who depend on it programmatically.
Olsen was an early public advocate for incorporating language models into that workflow. His March 2023 roadmap distinguished a launched Query Explanations feature from proposed translations, natural-language querying, improved search and a knowledge assistant. Keeping those stages separate is essential; a roadmap documents intent rather than proving universal delivery. The underlying challenge he identified was especially practical: generating plausible SQL is not enough if a system selects the wrong table from a large catalog.
By March 2026, Dune was announcing concrete additional interfaces for agents. Its CLI and Skills release described terminal-based dataset discovery, query execution and machine-readable results, alongside the previously launched MCP server. These are team product releases within Olsen's technical remit, not a claim that he individually designed each component. They extend the audience for Dune's data from people interacting with dashboards to software performing parts of the analytical workflow.
His record also includes a product that was discontinued. Olsen introduced Sim in May 2025 as a real-time data service extending Dune's ambitions beyond analytics, following the earlier Echo beta. In May 2026, he announced its scheduled August 1 closure and a renewed emphasis on the core Dune platform. Sim's website now redirects to the closure notice. Together, the announcements document both the initial rationale and the later narrowing of focus. They should be read together rather than presenting an earlier expansion plan as a permanent feature of the business.
Why Katharos.ai Follows Mats Olsen
Katharos.ai follows Olsen because the technical decisions behind a data platform influence the evidence available to everyone using it. Query engines, reusable models and access interfaces determine how much work an analyst must do before even beginning an investigation. When those components change, existing research can be affected as well as future applications. His public explanations provide a way to examine those choices beyond a company's marketing claims.
There is also a valuable tension between accessibility and analytical responsibility. Easier querying can widen participation, but it cannot determine whether an address represents a person, whether a transfer has economic significance or whether a comparison uses consistent definitions. The appropriate test of Olsen's work is therefore not simply whether an answer arrives quickly. It is whether the platform helps users inspect, reproduce and challenge the steps that produced it.
Areas to Follow
- Query correctness and continuity: numerical precision, changing schemas, migration support and reproducibility across versions of the data platform. - Community data engineering: contributor workflows, testing and ownership of reusable models, with credit preserved for their actual authors. - AI interfaces: table discovery, visible SQL, error handling and the limits placed on automated interpretation of blockchain activity. - Sustainable product scope: which services receive long-term support after Sim's announced closure, and how clearly customers are informed when priorities change.
Reporting Scope
This profile covers Olsen's public CTO role, documented product communications and Dune's relevant engineering milestones. It distinguishes his own explanations from work performed by Dune teams, upstream open-source projects and community analysts. Supplier case studies are identified as such, and historical launch plans are not assumed to describe the current service. It does not infer individual coding authorship, private investments or personal wealth from an executive title.
Areas of Influence
Olsen belongs in Exchanges & Infrastructure through data-platform engineering, Investment & Founders through co-founding Dune, and Community, Media & Education through the tools and explanations supporting collaborative analysis. His influence is strongest where open blockchain records must become usable evidence or dependable inputs for other software. That includes both enabling new forms of research and making the limitations of the data-processing system understandable.
Katharos.ai publishes this profile for information, based on the sources listed. Inclusion is not an endorsement of the individual, their organizations, or any investment product.
Source notice: Information and images are attributed to the external sources listed. Katharos.ai does not endorse the subjects or claims described and does not independently certify the authenticity of every source statement or image.