--- title: "How Eye Security Built AI-Native Data Governance | Dawiso" description: "How Eye Security connected Claude to its data catalog over MCP, built a custom Metabase ingestion, and defined a business glossary across nine domains." url: "https://www.dawiso.com/case-study/eye-security" language: "en" --- [Eye Security](https://eye.security/) # AI-Native Data Governance for a Growing Cybersecurity Company Eye Security implemented Dawiso to connect Claude to its data catalog through MCP, build a custom Metabase ingestion, and define a business glossary across nine domains - laying the foundation for AI-grounded data governance at scale. ## The Client Eye Security is one of Europe’s fastest-growing cybersecurity platforms, founded in 2020 in The Hague by three former officers of the Dutch intelligence services (AIVD and MIVD). The platform combines AI-driven threat detection with a 24/7 SOC, incident response, and cyber insurance, giving organisations enterprise-grade security built on European principles and accountability. In June 2026, Eye Security raised a €60 million Series C led by Belgian investment company Sofina, with continued backing from existing investors TIN Capital and J.P. Morgan Growth Equity Partners. ​ ## The Challenge After several years of fast growth, Eye Security needed shared definitions and clear data ownership across its expanding business. Terms like active customer, churn, upsell, and endpoint carried different meanings in different teams. Definitions lived in Metabase descriptions and in individual analysts’ heads, without an agreed source. At the same time, Eye Security wanted to use LLMs more across the company - but only if the LLMs could answer questions using their own definitions and data, not from generic guesses. They were looking for a data catalog that could also serve as a context layer for AI. ​ > “Before we implemented Dawiso, the terms used to track business metrics on a day-to-day basis were not being documented consistently. Two stakeholders could have different ideas about the same number, with neither of them being wrong; they just had different definitions. During a company hack-week, a first version of a business glossary was created through a Confluence page, and it was only then that we realized how big the issue was, and that we needed a data governance tool to overcome this challenge.” Jesper Bergeskans, Head of Data Engineering, Eye Security ​ ## The Solution After evaluating several catalog tools, Eye Security chose Dawiso for two reasons: its AI context layer and a price-to-capability ratio that outperformed vendors offering similar functionality. The proof of concept started in August 2025 and moved fast. Together, we defined Spaces in Dawiso that mirror Eye Security’s business domains - Finance, Sales, Operations, SecOps, Product, Marketing, and a Company-wide layer for shared terms. Finance was built out first as the reference model, and the business glossary then expanded domain by domain. Because Metabase was Eye Security’s reporting layer, the team needed it brought into Dawiso properly. We built a custom Metabase ingestion application that connected the existing field and column descriptions in Metabase automatically to the matching glossary terms in Dawiso, so the documentation Eye Security analysts had already written started paying off without anyone redoing it. Snowflake and dbt were scanned and catalogued alongside Metabase, and from day one of the PoC, Dawiso’s MCP server was wired into Claude, making the catalog accessible through chat as well as through the Dawiso UI. ​ #### **Dawiso: A Single Workspace for Data and Knowledge Management** Dawiso is a single workspace that elevates data management maturity, knowledge management, and collaboration. By linking business semantics with technical metadata, it serves as a unified system for information and metadata management, catering to all teams and user roles. ​ ### **Main features of Dawiso solution for Eye Security:** #### Unified Metadata Hub Information is available at one place regardless of the source of origin of the data. Dawiso brings together the company’s business dictionary, documentation, and technical metadata from a range of platforms deployed across the organizational structure. ​ #### Business Glossary by Domain Spaces in Dawiso reflect how a company divides responsibility. Each business domain has its own owners, terms, and access scope, with cross-domain definitions kept in a shared layer for the whole company to use. ​ #### AI Context Layer through MCP Dawiso’s MCP server gives Claude grounded access to the catalog. Glossary lookups, metadata queries, and lineage exploration all happen through chat - providing a reliable context layer for any AI work the customer builds on top. ​ #### Custom Connectors When Needed When standard connectors are not enough, the Dawiso team builds custom ingestion applications that preserve the customer’s existing documentation work and integrate tightly with their reporting stack. ​ #### Quick and Easy Implementation Dawiso seamlessly integrates into your existing environment. Its scanners are compatible with a wide array of data and analytics platforms, ensuring that prior solutions are built upon, and no data is lost in the transition. ​ ## The Impact Eye Security now has a governance foundation across nine business domains, with more than 40,500 data objects scanned across Metabase, Snowflake, and dbt. AI-generated descriptions enriched the technical layer in hours, not months. The MCP integration with Claude is live, and the team is already preparing the next step - an internal Slack-based AI assistant grounded in the Dawiso catalog, so any analyst can ask where a number comes from and get a governed answer in the tools they already use. With Snowflake, dbt, and Metabase all visible in one place, the data team has, for the first time, a complete view of how data flows from warehouse through transformation to reporting. ​ > “As the world enters a new age of AI, this also changes how we find information. Google has integrated AI Overviews into its search results, and questions regarding specific topics have moved away from message board forums and into LLMs. By connecting our internal AI platform to Dawiso’s MCP server, we give our business users the same flexibility and clarity. If they search for a general term, one with several sub-definitions, we can present all of those to them in an easy-to-consume way. It’s a game-changer for how we communicate internally.” Jesper Bergeskans, Head of Data Engineering, Eye Security ​ ## Next Phase Eye Security is now moving toward end-to-end lineage across dbt, Snowflake, and Metabase, expanding the business glossary into SecOps (including MITRE ATT&CK terminology and SOC workflows), and connecting Dawiso to BI tools through the Database API to give leadership real-time visibility into governance health and data flow. ## Key Numbers - 9 governed business domains - 40,500+ scanned objects across Metabase, Snowflake, and dbt - 12,500+ Snowflake objects - 28,000+ dbt objects from the Analytics project - 3,200+ Metabase reports and dashboards connected via a custom ingestion application - MCP server wired into Claude from day one of the proof of concept Martin Sklenička Data Governance Consultant [Dawiso on LinkedIn](https://cz.linkedin.com/company/dawiso) ## More Case Studies ### [300+ Data Products Someone Is Responsible For](https://www.dawiso.com/case-study/kb-data-products) Komerční banka turned a 370,000+ table data estate into 300+ data products in Dawiso, each with a named owner, business lineage between products and confirmed consumers. Banking 300+ data products Read more ### [A governed knowledge foundation for people, applications and AI](https://www.dawiso.com/case-study/raiffeisenbank) Raiffeisenbank moved around 4,000 internal standards out of SharePoint into Dawiso, with a full lifecycle and approval model, point-in-time reconstruction, and REST API and MCP access. Banking ~4,000 governed documents Read more ### [Building the foundation for data governance](https://www.dawiso.com/case-study/olvi) Finnish beverage group Olvi defines and maintains business terms in Dawiso across 9 thematic spaces: 860+ terms, 80 active users, and 39,000+ documented objects. Food & Beverage 39,000+ documented objects Read more ### [From open-source DataHub to an AI-ready Dawiso catalog in two months](https://www.dawiso.com/case-study/bertel-o-steen) Bertel O. Steen migrated 100,000+ Databricks objects, 300+ bilingual business terms and 50+ data products from DataHub to Dawiso, with end-to-end lineage and a Databricks MCP integration. Automotive & Retail 100,000+ scanned objects Read more ### [A trusted AI context layer for Seznam.cz](https://www.dawiso.com/case-study/seznam) One trusted AI context layer over MCP, with end-to-end lineage across Keboola, Snowflake and Tableau and governed data products. Technology & Media 681,000 governed objects Read more ### [Support for data-driven company management](https://www.dawiso.com/case-study/stora-enso) Connecting It All: Power BI, Wherescape, Snowflake, Azure Data Factory, SQL, and the Entire Data Community Meeting in Dawiso. Manufacturing 8,700+ Power BI reports scanned Read more ### [Unlocking Seamless Data Management Through Dataportal](https://www.dawiso.com/case-study/kooperativa) Over 3.4 Million Objects Available for Everyday Work of More Than 5,900 Users. Insurance 3.4M+ objects in the environment Read more ### [One Hub Connecting 30+ Financial Institutions](https://www.dawiso.com/case-study/kb) Simplifying Work with Data and Reports in a Large Enterprise. Banking 370,000+ data tables Read more ### [From Scratch to Comprehensive Data Governance in 78 Days](https://www.dawiso.com/case-study/p3-parks) A seamless transition from spreadsheets to a user-friendly data governance platform. Real Estate 78 days to full rollout Read more