AI-Powered Data Catalog

Enterprise Data Catalog Built for Scale

Your pipelines pass. Your dashboards break anyway. DataHub gives platform engineers unified metadata, AI anomaly detection, and column-level lineage across 87+ sources.

  • Semantic search finds datasets without exact keyword matches

  • ML anomaly detection catches volume and freshness issues early

  • Column-level lineage maps every upstream dependency

Explore the product

See DataHub in your environment

Request a scoped demo with a DataHub engineer.

Trusted by modern data teams
The real problem

What does data catalog failure actually cost?

Bad metadata compounds. One undocumented table becomes a broken dashboard, a missed SLA, and an audit you cannot answer.

Discovery debt

Engineers spend hours finding the right dataset, slowing delivery and increasing the cost of every data request.

Silent pipeline failures

Bad data reaches production undetected, eroding trust in dashboards and triggering costly incident reviews.

Governance gaps

No audit trail when compliance asks who owns what, leaving teams exposed during reviews and regulatory inquiries.

Lineage blind spots

Downstream impact stays unknown until something breaks, turning routine changes into unplanned incidents.

The solution

A modern data catalog built for your stack

DataHub connects to your existing infrastructure and gives every team the metadata context they need to move faster and govern confidently.

AI-powered search and discovery

DataHub indexes metadata from 87+ sources and surfaces the right dataset using semantic understanding, not just keyword matching.

  • Natural language queries return ranked, relevant results
  • Descriptions and tags generated from existing metadata
  • Search scope filtered by domain, owner, or data tier

Proactive AI data anomaly detection

ML models monitor volume, schema, and freshness across your pipelines and surface deviations before they reach downstream consumers.

  • Volume and freshness checks run on a defined schedule
  • Alerts route to Slack, PagerDuty, or your incident tool
  • Anomaly history tracked per dataset for trend review

Column-level lineage and impact analysis

Trace data from source to dashboard at the column level. Understand which pipelines, models, and reports are affected before you make a change.

  • Column-level lineage across SQL, dbt, and Spark jobs
  • Impact analysis shows affected dashboards and models
  • Lineage graph exportable for audit and review purposes

Fine-grained access control and governance

Define ownership, apply data classifications, and enforce access policies at the domain or dataset level without rebuilding your existing stack.

  • Role-based access policies applied at the domain level
  • Ownership and stewardship assigned per dataset or schema
  • Audit logs capture every metadata change with attribution
Getting started

How it works

Three steps from connection to full catalog governance. No pipeline rebuilds required.

01. Connect your sources

  • Ingest from databases, warehouses, and lakes
  • Pre-built connectors for Snowflake, dbt, Airflow, and more
  • Custom sources added via the open ingestion framework

02. Contextualize your assets

  • Ownership and classifications applied automatically
  • AI-generated descriptions refined by your team over time
  • Glossary terms linked to datasets across every domain

03. Activate governance platform-wide

  • Search, lineage, and anomaly alerts available to every team
  • Access policies enforced at the domain or dataset level
  • Audit logs and compliance reports generated on demand
Enterprise readiness

Built for enterprise-grade cloud data governance and catalog

DataHub fits the infrastructure model your security and platform teams already operate. No rearchitecting required.

Flexible deployment options

  • Deploy on your cloud, in a private VPC, or as managed SaaS
  • Kubernetes-native with Helm chart support
  • Fits your existing infrastructure model without migration

Security and compliance controls

  • SSO via SAML and OIDC with RBAC at domain level
  • Full audit logging for every metadata change
  • Data classification supports SOC 2 and GDPR requirements

Open API and extensibility

  • GraphQL and REST APIs for custom integrations
  • Embed metadata into internal tools and pipelines
  • Apache 2.0 licensed with 12K+ GitHub stars
Customer voices

Trusted by modern data teams

Platform engineers and data leaders across financial services, healthcare, and technology rely on DataHub to govern their data platforms.

Gartner Peer Insights
Verified review
Key outcome
Lineage visibility
"DataHub gave us the lineage visibility we needed to stop guessing about downstream impact when schemas change."
DR
Verified Reviewer
Data Platform Engineer, Enterprise Technology
Gartner Peer Insights
Verified review
Key outcome
Self-service search adoption
"The search experience is the first one our analysts actually use without asking the data engineering team for help."
MK
Verified Reviewer
Senior Data Engineer, Financial Services
Gartner Peer Insights
Verified review
Key outcome
Automated compliance history
"We passed our SOC 2 audit with a complete ownership and access history that DataHub maintained automatically."
SL
Verified Reviewer
Head of Data Governance, Healthcare

Frequently asked questions about enterprise data catalog

Most teams complete initial ingestion and search configuration within two weeks. Full governance workflows, including ownership and classification, typically follow in the next sprint cycle. The timeline depends on the number of sources you connect and how much metadata enrichment your team wants to apply upfront.
DataHub supports SSO via SAML and OIDC, role-based access control at the domain and dataset level, and full audit logging. It does not store your underlying data, only metadata. Access policies are enforced within DataHub and do not require changes to your source systems.
DataHub ships with connectors for 87+ sources, including Snowflake, BigQuery, Redshift, dbt, Airflow, Looker, Tableau, and Kafka. Custom sources can be added via the open ingestion framework. The connector library is maintained by both the DataHub team and the open source community.
DataHub parses SQL, dbt models, and Spark jobs to extract field-level relationships. You can trace any column from its origin through every transformation to its final destination in a report or model. The lineage graph is queryable via API and exportable for audit purposes.
DataHub includes AI-assisted metadata generation for descriptions and tags, semantic search using vector embeddings, and ML-based anomaly detection for volume, freshness, and schema drift. These features work on top of your existing metadata and do not require separate model training or data exports.

Ready to bring order to your data platform?

Talk to a DataHub engineer about your environment. We will scope a demo around the problems your team is actually facing.

Request a demo Explore the product
Apache 2.0 open source
60+ pre-built connectors
Self-hosted or managed deployment