DataHub Cloud

DataHub Cloud data catalog

DataHub Cloud adds certification, structured properties, propagation automations and a data product marketplace to the open-source catalog.

Compare: data catalog tools

• Data Catalog: Core product with 100+ integrations for automated metadata harvesting, AI-powered search, business glossary, and usage-aware search ranking (docs).

• Automated Metadata Harvesting: 100+ source integrations automatically ingest schemas, tables, columns, and pipeline metadata; remote ingestion agent supports on-premises sources (docs).

• Search & Discovery: Enhanced usage-aware search ranking surfaces the most relevant assets first; faceted filtering across all catalog dimensions (docs).

• AI-Powered Search: Ask DataHub AI Agent enables conversational discovery — users ask plain-language questions and get trusted data results and SQL generation (website).

• Usage & Popularity Statistics: Usage-aware search ranking is derived from actual query usage statistics per asset — surfaces most-used assets higher in results (website).

• Custom Fields & Templates: Structured Properties add typed custom fields to any asset, in Core and Cloud (docs).

• Asset Certification: Certification marks trusted assets (public beta, Context add-on) (docs).

• Catalog Usage Analytics: Analytics track usage and documentation progress, including compliance-form analytics (docs).

• Business Glossary: Business glossary with term creation, asset linking, and owner-based review/approval workflows, plus related-term and parent/child hierarchy relationships in the metadata graph — accessible via MCP (docs).

• Metadata Propagation: Automations propagate documentation, glossary terms and domains along lineage (docs).

DataHub Cloud data lineage

DataHub Cloud shows table, column and end-to-end lineage with impact analysis, and accepts OpenLineage events.

Compare: data lineage tools

• Data Lineage: Column-level and table-level lineage across 100+ data sources; lineage accessible via MCP and visible to AI agents (website).

• Table-Level Lineage: Dataset-to-dataset lineage tracking across warehouses, ETL tools, and BI platforms (website).

• Column-Level Lineage: Column-level lineage tracking field transformations from source to destination — accessible via MCP tools (docs).

• End-to-End Lineage: End-to-end lineage from source systems through transformations to downstream AI models and BI dashboards confirmed (website).

• Impact Analysis: Lineage graph enables impact analysis — identifying all downstream consumers before making upstream changes (website).

• OpenLineage Support: DataHub is an OpenLineage backend — natively consumes OpenLineage events from Airflow, Spark, dbt, and other emitters (docs).

DataHub Cloud data stewardship

Compare: data stewardship tools

• Data Stewardship: Custom ownership types (e.g. data steward, business owner) are assigned to assets, and owners review change proposals in the Task Center (docs).

• Review & Approval Workflows: Approval workflows for documentation, tags, glossary terms, and access requests are a confirmed DataHub Cloud feature (docs).

• Governance Campaigns: Compliance Forms assign questions (owners, descriptions, terms, domains, properties) across a set of assets to owners or ownership types, notify them in their inbox, and track completion in analytics (docs).

• Metadata Automation: Workflow engine with metadata tests and approval workflows automates governance actions — documentation, tags, glossary terms, and access requests (docs).

• Critical Data Elements: Not explicitly confirmed as a dedicated CDE feature (docs).

• Governance Scorecards: Not confirmed in product pages reviewed (website).

DataHub Cloud data classification

Compare: data classification tools

• Sensitive Data Discovery: Classification tags can be applied to assets, but automated PII/PHI scanning via ML classifiers was not confirmed for DataHub Cloud (docs).

• Data Policy: Fine-grained access control exists at the catalog level, but data-level policy enforcement (row/column security at source) was not confirmed (docs).

• Data Masking: Not confirmed as a DataHub Cloud feature (docs).

• Native Policy Enforcement: Snowflake, BigQuery and Databricks sync automations push tags and descriptions, not access policies (docs).

DataHub Cloud AI features

Ask DataHub answers questions about your data, and DataHub Cloud adds custom agents, context evals and an agent registry.

• AI Features: Ask DataHub AI Agent, Observe AI Agent, AI documentation generation, AI anomaly detection, and hosted MCP Server all confirmed (website).

• AI Chat Assistant: Ask DataHub chat agent enables conversational data discovery — finding trusted metrics, generating SQL, and debugging quality issues in natural language (website).

• AI Documentation Generation: AI documentation generation and intelligent glossary classification — lineage-based auto-enrichment of catalog assets at ingestion time (website).

• Agentic Metadata Enrichment: Observe AI Agent for automated check setup and health reporting; AI agents for catalog maintenance via MCP server — autonomous metadata enrichment confirmed (website).

DataHub Cloud data compliance

• Audit Trail: Audit trails tracking access, metadata changes, and governance actions are confirmed for regulatory evidence (website).

• Data Compliance: Compliance forms are mentioned but detailed audit trail and regulatory reporting capabilities were not explicitly confirmed (website).

• Compliance Reporting: Not explicitly confirmed as a dedicated compliance reporting feature (website).

• Regulatory Framework Mapping: Not confirmed in product pages reviewed (website).

• Data Activity Monitoring: Not confirmed; source-level data access monitoring was not found (website).

• ROPA: Not confirmed; no explicit GDPR Article 30 processing register found (website).

• Data Retention Policies: Not confirmed in product pages reviewed (website).

• Data Subject Rights: No DSAR automation or right-to-be-forgotten workflows confirmed (website).

DataHub Cloud collaboration

• Collaboration: Slack integration with Ask DataHub bot, subscriptions and notifications, and approval workflow notifications all confirmed (docs).

• Alert Notifications: Slack integration with bot commands, subscription-based notifications, and quality alert routing to channels and DMs confirmed (docs).

• Discussions & Annotations: Inline conversations and annotations on catalog assets were not explicitly confirmed (docs).

• Ticketing Integrations: Jira or ServiceNow integrations were not confirmed in pages reviewed (docs).

DataHub Cloud data marketplace

• Self-Service Access Requests: Approval workflows for access requests are a confirmed DataHub Cloud feature — consumers request access and requests route to data owners (docs).

• Data Products: Data products group assets with owners, and v2.2.0 added a Data Product Marketplace page (release notes).

• Data Mesh Support: DataHub Cloud supports data mesh patterns with data domains, first-class data products, and federated governance across domain owners (docs).

• Subscription Management: Subscription and Notification Management is a dedicated DataHub Cloud feature — users subscribe to data assets and receive change notifications (docs).

• Data Marketplace: No dedicated self-service data marketplace storefront confirmed in DataHub Cloud (website).

DataHub Cloud data quality

Compare: data contracts tools

• Data Quality: Observe AI Agent with automated anomaly detection, freshness/volume/schema monitoring, custom SQL checks, and a Data Health Dashboard (website).

• SQL Table Tests: Custom SQL checks for freshness, completeness, and custom assertions are part of the DataHub Cloud observability feature set (website).

• Data Health Scores: Data Health Dashboard aggregates quality signals into per-asset health status with real-time quality alerts (website).

• Anomaly Detection: AI anomaly detection automatically spots unusual patterns in data volume, freshness, and schema without manually defined thresholds (website).

• Data Profiling: SQL profiling during ingestion computes table and column statistics (docs).

• Data Contracts: Data contracts bundle schema, freshness, and data quality assertions on a dataset, created from the UI (Quality > Data Contracts tab) or the API. Listed as available in both DataHub Core and DataHub Cloud (docs).

DataHub Cloud AI governance

• AI Governance: The metadata model has first-class MLModel, MLModelGroup and MLFeature entities, so models are cataloged and governed alongside datasets rather than tracked separately (docs).

• AI Asset Inventory: MLModel entities carry versions, owners, hyperparameters, training metrics and evaluation data, giving a queryable registry of models in the catalog (docs).

• AI Lineage: MLModel aspects record the training data used to produce a model, linking models back to the datasets that trained them in the lineage graph (docs).

• AI Agent Access Governance: Enforces fine-grained access controls at the context layer: policies govern which agents and service accounts can access which datasets, columns, business terms, and sensitivity-classified fields, and the same policy engine that governs human access applies to every agent request. Every context access is logged with provenance metadata (website).

• AI Regulatory Compliance: No EU AI Act, NIST AI RMF or ISO 42001 assessment templates or evidence workflows — model metadata only (docs).

• AI Model Health Score: No composite model trust or readiness score aggregating governance signals (docs).

DataHub Cloud workflows

• Workflows: API, MCP server, Python SDK, dbt integration, OpenLineage, Chrome extension, and Slack integration all confirmed (docs).

• REST / GraphQL API: DataHub API (REST and GraphQL) for programmatic access to all catalog metadata — documented in the managed DataHub API section (docs).

• MCP Server: DataHub hosted MCP Server is a confirmed Cloud feature — connects AI tools to the catalog for asset lookup, lineage, tags, glossary, and ownership (website).

• dbt Integration: dbt connector is one of the most popular DataHub integrations — ingests models, sources, tests, and column-level lineage from dbt Cloud and Core (docs).

• Python SDK: acryl-datahub Python package (pip install acryl-datahub) is the primary SDK for programmatic catalog interaction (docs).

• Browser Extension: Chrome extension surfaces DataHub metadata within Looker, Tableau, Power BI, BigQuery, Databricks, Mode Analytics, Superset, Hex, and Metabase (docs).

• Webhooks: Not confirmed in DataHub Cloud documentation reviewed (docs).

DataHub Cloud data context layer

• Knowledge Graph: DataHub's core architecture is a metadata graph — all assets, users, groups, tags, terms, and policies are nodes with typed edges, queryable via GraphQL (docs).

• Semantic Models: Metrics and semantic models are catalogued as first-class assets (beta, v2.1.0) (docs).