Metadata Management

Business Glossary

Model business meaning as an ontology so teams share one vocabulary, and AI assistants ground discovery and access in concepts your organization already trusts.

business-glossary-ontology-gif.gif

Overview

Blindata’s Business Glossary is your shared ontology: concepts, relationships, and namespaces that answer what this data means. The Blindata AI Assistant and connected agents use that map to reach the right data products and assets, instead of relying on keyword matches alone.

People and agents share the same path from meaning to data. When a person searches for “Customer” or “Order,” or an agent resolves those concepts, Blindata navigates the ontology first, then follows semantic links to certified data products and physical assets—so discovery, definitions, suggested links, and queries against mapped tables and columns stay grounded in stewarded vocabulary, under the same access controls as the rest of the platform.

Features

Build a shared ontology that teams and agents can trust. Domain-driven namespaces keep ownership and scope clear, relationships connect concepts, and semantic links bind them to real catalog assets and data products. Use the Blindata AI Assistant to draft and refine a map of your data that humans and agents can consume.

Domains & namespaces

Ontology is domain-driven: each namespace is a business domain with clear ownership and scope, so Customer in Sales is not confused with Customer in Support.

Relationships

Model how the business actually works with inheritance, composition, and dependencies, so people and agents can walk the graph instead of skimming a flat term list.

Semantic links

Bind glossary attributes to catalog columns and data product ports so every concept points at certified assets, not just a definition.

Stewarded term detail

Aliases, formulas, patterns, naming rules, and taxonomies live on the term, so stewards keep precision where it belongs without bloating the graph.

AI-assisted curation

Draft concepts and structure from docs, diagrams, or catalog assets with the Assistant. Stewards review and save so the ontology stays trusted.

Grounded AI context

Assistants and agents follow the same path from namespace to concept to links to schema, so answers and queries land on meaning your organization already owns.

How to

Building the ontology is an AI-assisted process. Use the Blindata AI Assistant to draft concepts from material you already have: attach policy PDFs or specifications, upload pictures or data-model diagrams, or start from the Data Catalog (tables, columns, and descriptions that suggest business terms). Review the proposed concepts, attributes, and namespaces in Blindata, then refine ownership and definitions so the vocabulary is stewarded, not only generated.

Use the graph editor, and the Assistant when you need a first draft, to link concepts the way the business connects them: inheritance, composition, and domain relationships. Ask the Assistant to propose structure from an attached diagram or document, then adjust in the editor until the ontology matches how the business works. Those relationships are what people and agents follow from a business question to the right data.

Semantically link glossary attributes to catalog columns and data product ports so meaning is bound to real assets. The Assistant can suggest links between concepts and technical metadata; you stay the reviewer. Producers keep links current as products evolve. Without them, discovery and AI fall back to keyword search.

Once concepts, relationships, and semantic links are in place, search and the Assistant navigate from ontology to data products to schema. Ask for a business concept and get the certified assets and mappings your organization has already agreed on. The same path powers the Context Layer for external agents, with your access controls intact.