OpenMetadata

OpenMetadata is the leading open-source context layer for data cataloging, metadata management, and AI agent integration, built by the original founders of Apache Hadoop, Apache Atlas, and Uber Databook. The platform unifies technical context (schemas, tables, columns, dashboards, pipelines, ML models, owners, lineage, classifications), business semantics (glossary terms, metric definitions, ontologies, W3C-standard RDF, OWL, DCAT, Schema.org relationships), and organizational memory (corrections, decisions, approvals, feedback loops, auditable change records) into a single Semantic Context Graph that both humans and AI agents can reason from. With over 4,000 enterprise deployments, 13,500 open-source community members, 450 code contributors, and 14,000 GitHub stars, OpenMetadata is the fastest-growing open-source project in the data catalog space. The platform features an API-first and schema-first architecture with over 700 open specifications in JSON Schema and RDF/JSON-LD, making metadata fully programmable and extensible. OpenMetadata provides 130+ data connectors covering databases (PostgreSQL, MySQL, Snowflake, BigQuery, Redshift, Databricks, ClickHouse, MongoDB), messaging systems (Kafka, Redpanda, Kinesis), dashboards (Tableau, Looker, Power BI, Metabase, Superset), pipelines (Airflow, Dagster, dbt, Fivetran), ML models (MLflow, SageMaker), and storage systems. The streamlined architecture has only four system components for simplified deployment, and includes a native MCP server for direct agent-to-context integration and an AI SDK for programmatic access. Trusted by OpenAI, Unity, Airbus, Yelp, Scout24, Rakuten, Union Bank, and Carrefour Brazil, OpenMetadata manages 2+ million data assets at large deployments.