Dawiso is an enterprise data catalogue with an AI context layer on top. It scans metadata across more than 40 data platforms and enriches it automatically, holds a business glossary so terms mean one thing across an organisation, draws interactive lineage showing where a number came from, and adds governance for unstructured data and for AI use itself. The company is Dawiso s.r.o., and it names Société Générale, Kooperativa and ČEZ Group among 24,000-plus data professionals using it.
Exposing the catalogue to AI agents over the Model Context Protocol is the forward-looking part, and it is more consequential than it first sounds. An agent asked a question about company data has the same problem a new analyst does — it does not know which of four tables called `revenue` is the trusted one — and a governed glossary and lineage graph is exactly the context that answers it. Most catalogues were built to be read by people; wiring one to be read by agents is where this is heading. Named enterprise clients of that size are also stronger evidence than any feature list.
Two things to weigh. ‘Over 50% less than other solutions’ is a comparison with no number attached, and enterprise data catalogues are sold on quotes, so the actual figure requires a call. And the honest risk in this category is not the software: catalogues fail on adoption rather than features, because a glossary nobody maintains becomes wrong faster than having none, and automated enrichment reduces that work without removing it. Budget for the stewardship, not just the licence.
