OpenMetadata

The Open Semantic Context Platform for Data and AI

OpenMetadata is the open platform for building trusted data context and business semantics for humans, AI assistants, and agents.

OpenMetadata connects technical metadata, data quality signals, data lineage, column-level lineage, ownership, usage, policies, conversations, glossaries, classifications, metrics, domains, and data products into a unified metadata knowledge graph. With 120+ connectors, open metadata standards, semantic search, APIs, SDKs, and an MCP server, OpenMetadata gives every user and AI system the governed context it needs to discover, understand, trust, and use data.

AI does not need another raw database connector. AI needs context.

OpenMetadata provides that context:

• what data exists
• what it means
• who owns it
• how it is used
• where it came from
• where it flows
• whether it is fresh, tested, and trusted
• which business concepts, glossary terms, classifications, and policies apply
• what downstream assets, dashboards, pipelines, metrics, and ML models depend on it

Contents

• Why OpenMetadata for AI?
• Context: Give AI the Full Picture of Your Data
• Semantics: Give AI Business Meaning
• Knowledge Graphs and Ontologies
• Automation: Activate Context and Semantics with AI
• What You Can Build
• How OpenMetadata Works
• MCP: Connect AI Assistants and Agents
• Semantic Search
• OpenMetadata Standards
• Core Platform Capabilities
• Quickstart
• Documentation and Community
• Contributing
• License

Why OpenMetadata for AI?

AI needs more than data access. It needs context, semantics, trust, lineage, governance, and operational awareness.

Connecting an AI assistant directly to a database, warehouse, dashboard, or pipeline only gives it raw access to data structures. It does not give the AI enough context to understand what the data means, whether it can be trusted, who owns it, how it is governed, or what downstream systems depend on it.

OpenMetadata gives AI systems the context and semantics they need to safely discover, understand, govern, and use enterprise data.

OpenMetadata does this by combining four capabilities:

1. Context — technical, operational, trust, and lineage metadata from the data ecosystem.
2. Semantics — business meaning through glossaries, metrics, classifications, domains, policies, and ontologies.
3. Knowledge Graph — relationships connecting assets, columns, people, teams, policies, lineage, quality, and business concepts.
4. Automation — MCP, Semantic Search, APIs, SDKs, events, and workflows that let AI assistants and agents act on governed metadata.

With OpenMetadata, AI can answer questions such as:

• What does this metric mean?
• Which datasets power this dashboard?
• Who owns this data product?
• Is this dataset certified, fresh, and high quality?
• What downstream dashboards or ML models are affected by this column change?
• Which assets are related to customer purchase behavior, even if they use different names?
• Which columns contain sensitive customer information?
• Which glossary terms and business concepts apply to this dataset?

Context: Give AI the Full Picture of Your Data

Context is the metadata that describes how data exists, behaves, changes, flows, and is used across the organization.

OpenMetadata collects context from across your data stack and connects it into a unified metadata graph.

Technical Metadata

OpenMetadata gives AI access to technical metadata such as:

• databases, schemas, tables, columns, topics, dashboards, charts, pipelines, APIs, search indexes, ML models, and storage assets
• schemas, column names, data types, constraints, descriptions, sample queries, joins, and service metadata
• service configuration, ingestion metadata, and operational metadata
• owners, teams, users, personas, domains, data products, and usage patterns

Data Quality and Trust Signals

AI should not treat every dataset as equally trustworthy.

OpenMetadata gives AI access to trust signals such as:

• data quality tests
• test suites and test results
• freshness checks
• volume checks
• null, uniqueness, distribution, and custom tests
• profiling results
• observability signals
• data quality history
• incidents, alerts, and operational health signals

Data Lineage and Impact

AI needs to understand where data comes from and where it goes.

OpenMetadata captures:

• upstream and downstream lineage
• table-level lineage
• dashboard lineage
• pipeline lineage
• metric lineage
• ML model lineage
• API and topic dependencies
• impact analysis across the data estate

Column-Level Lineage

For precise AI reasoning, table-level lineage is not enough.

OpenMetadata helps AI understand:

• which source columns produce which downstream columns
• how columns flow through transformations
• which dashboards, reports, metrics, or ML models depend on a specific column
• what may break when a column changes

Connected from 120+ Data Services

OpenMetadata brings this context together from databases, warehouses, lakes, dashboards, pipelines, messaging systems, ML platforms, storage systems, APIs, search systems, and metadata systems.

Context answers questions like:

• What data exists?
• Where did this data come from?
• Who owns it?
• Is it fresh?
• Is it tested?
• Is it trusted?
• What systems depend on it?
• What happens if it changes?

Semantics: Give AI Business Meaning

Semantics is the business meaning layered on top of technical context.

Without semantics, AI may see a column named cust_id, acct_id, or buyer_key, but it may not know whether those fields represent a customer, an account, a buyer, a household, or a legal entity.

OpenMetadata lets teams define, govern, and connect business meaning across the metadata graph.

Business Concepts

Define the concepts that matter to the business, such as:

• Customer
• Account
• Order
• Revenue
• Product
• Consent
• Churn
• Risk
• Lifetime Value
• Net Retention
• Active User
• Sensitive Data

Glossaries and Glossary Terms

OpenMetadata lets teams create governed vocabularies with:

• business definitions
• synonyms and abbreviations
• owners and reviewers
• related terms
• hierarchical terms
• links to tables, columns, dashboards, metrics, and data products

Metrics and KPIs

Metrics are one of the most important semantic objects for AI.

OpenMetadata helps AI understand:

• what a metric means
• how it is calculated
• who owns it
• which dashboards use it
• which tables power it
• which glossary terms define it
• which downstream consumers depend on it

Classifications and Tags

OpenMetadata lets teams classify and label data with governed tags such as:

• PII
• Sensitive
• Confidential
• Certified
• Deprecated
• Tier 1
• Finance
• Marketing
• GDPR
• HIPAA
• SOX
• ML Feature
• Customer Data

Domains and Data Products

OpenMetadata connects assets to business ownership boundaries through:

• domains
• data products
• teams
• owners
• policies
• personas
• data product consumers

Policies and Governance

OpenMetadata connects semantics to governance so AI systems can reason with policy-aware context, not just metadata.

This includes:

• ownership
• stewardship
• classification
• access control context
• certification
• review workflows
• governance policies
• lifecycle states

Semantics answers questions like:

• What does this data mean?
• What business concept does this column represent?
• Is this metric officially defined?
• Is this asset certified?
• Is this data sensitive?
• Which glossary terms apply?
• Which domain owns this data product?

Knowledge Graphs and Ontologies

OpenMetadata connects context and semantics into a unified metadata knowledge graph.

The graph does not just store data assets. It stores the relationships between data assets, people, teams, policies, quality tests, lineage, classifications, glossary terms, metrics, domains, and data products.

This makes OpenMetadata a semantic context layer for AI.

Example relationships:

Table ──hasColumn────────────> Column
Column ──classifiedAs────────> PII
Column ──represents──────────> Customer Identifier
Table ──ownedBy──────────────> Data Engineering Team
Table ──partOf───────────────> Customer 360 Data Product
Dashboard ──dependsOn────────> Table
Metric ──definedBy───────────> Glossary Term
Pipeline ──produces──────────> Table
Column ──flowsTo─────────────> Column
Test Case ──validates────────> Table
Domain ──contains────────────> Data Product
Glossary Term ──relatedTo────> Business Concept
Policy ──governs─────────────> Classification

With this graph, AI can reason across relationships:

• Which datasets power this dashboard?
• What does this metric mean?
• Who owns this data product?
• Is this table fresh, certified, and high quality?
• Which downstream dashboards or ML models are affected by this