Awesome Data & AI Governance

This repository contains a curated list of awesome Data & AI Governance platforms and tools that help you discover, manage, and observe data and AI solutions in your organization.

Contents: Data Discovery and Observability Solutions

---

GenAI Readiness Features

As Generative AI (GenAI) solutions gain prominence, the role of robust Metadata Management Systems equipped with Large Language Model (LLM) capabilities becomes increasingly critical. Such systems significantly enhance context-awareness, enabling advanced semantic understanding, bridging the gap between business and technical metadata, and effectively powering Retrieval-Augmented Generation (RAG) and Agent-to-Agent (A2A) interactions. We observe a notable convergence where RAG systems, initially context-limited, incorporate Metadata Management capabilities, while Metadata Management platforms increasingly adopt GenAI functionalities to cater to new, sophisticated use-cases.

The following categories outline key GenAI readiness features available across various Metadata Management solutions:

• AI & Semantic Metadata Enrichment
• Semantic Translation & Business Glossaries
• Technical Metadata & Query Analytics
• Data Quality, Observability & Governance

Disclaimer: The GenAI feature landscape evolves rapidly, and thus the provided list, feature groupings, and matrix may frequently change. Contributions, updates, and community-driven insights are strongly encouraged to keep this resource comprehensive and current.

---

AI & Semantic Metadata Enrichment

The group includes built-in AI capabilities essential for Generative AI applications, enhancing metadata through semantic enrichment and context awareness. It supports Agent-to-Agent (A2A) protocols and Retrieval-Augmented Generation (RAG) systems by facilitating semantic understanding, vectorization, and natural language interactions. These capabilities align closely with Model Context Protocol (MCP), enabling standardized metadata exchange and interoperability across GenAI applications, making it possible for AI agents and RAG systems to leverage rich, structured metadata efficiently.

Semantic Translation & Business Glossaries

Features enabling a semantic bridge between business concepts and technical metadata. These capabilities enable precise mapping between business terms and data assets, fostering accurate, context-aware query generation.

Technical Metadata & Query Analytics

Core features providing detailed technical insights, query analytics, and usage patterns critical for metadata-informed query generation, optimization, and efficient GenAI-driven data interactions. Essential for automated database query generation, lineage tracking, and metadata-informed GenAI processes.

Data Quality, Observability & Governance

Essential features for maintaining data reliability, trustworthiness, observability, collaborative governance, and secure access management. Integral to ensuring metadata-driven AI operations remain accurate, dependable, securely managed, and compliant with sensitive data handling requirements.

High-Level Feature Comparison

Definitions:

• Specification-based - uses an open standard for collecting metadata to allow efficient time-to-discovery and federating data catalogs
• Search-based - allows to search for data assets
• Network-based - provides rich context about data asset ownership
• Lineage-based - provides lineage for all entities the solution operates
• Federation - the ability to map multiple data catalogs into a single UI to avoid repeated data collection.
• ML 1st citizen - operates ML entities on a high level - you can use them as any other data assets.
• Data Quality - includes mature data quality assurance tools.
• End-to-end lineage - data lineage that includes all data assets used in the organization across all its data catalogs and ML tools.
• Column-level lineage - data lineage with column level granularity
• Data collaboration - provides possibility to bring together data from various internal and external sources to unlock combined data insights

📙 Open-source Data & AI Governance Tools and Platforms

Amundsen

Website | GitHub

A popular open-source data catalog for metadata management and data discovery originated from Lyft. Created by Amundsen maintainers, Stemma provides a managed version of an enterprise data catalog, inspired by Amundsen.

More features

• Strategy: Push
• UX personalization: No
• AI autowiring: No
• Rich data profiling: No
• Recommendations: Yes
• Schemas, Description: Yes
• Complex schemas: No
• Data preview: Yes
• Column statistics: Yes
• Data owner: Yes
• Top data users: Yes
• Change notifications:No
• Change feed: No
• Deployment:
• Supported data sources: Hive, Redshift, Druid, RDBMS, Presto, Snowflake

DataHub

Website | GitHub

DataHub is an open-source data catalog enabling data discovery, data observability and federated governance that originated from LinkedIn and is commercially offered by Acryl Data as a cloud-hosted SaaS offering.

More features

• Strategy: Push, Pull
• Customizable metadata model: Yes. The metadata model can be declared using the open-source Pegasus language, and is interoperable with JSONSchema and Avro
• Rich data profiling: Yes
• Recommendations: Yes
• Schemas, Description: Yes
• Complex schemas: Yes
• Data preview: Yes
• Column statistics: Yes
• Data owner: Yes
• Top data users: Yes
• Lineage impact analysis: Yes
• Change notifications: Yes
• Change feed: No
• Automation: Yes
• UX personalization: No
• Deployment: docker-compose / Kubernetes with Helm, or using Acryl Data's SaaS offering
• Supported data sources:
  • Snowflake
  • BigQuery
  • Redshift
  • Hive
  • Athena
  • Postgres
  • MySQL
  • SQL server
  • Trino
  • Delta Lake
  • S3
  • Looker
  • PowerBI
  • Tableau
  • Mode
  • Metabase
  • Redash
  • Superset
  • Airflow
  • Great Expectation
  • dbt
  • Feast
  • SageMaker
  • Glue
  • Kafka
  • Nifi
  • Okta
  • LDAP
  • Slack
  • There's 50+ integrations - see the docs for the latest.

Marquez

Website | GitHub

Marquez is an open-source data catalog for collection, aggregation, and visualization of a data ecosystem’s metadata originated from WeWork.

More features

• Strategy: Push
• UX personalization: No
• AI autowiring: No
• Rich data profiling: No
• Recommendations: No
• Schemas, Description: Yes
• Complex schemas: No
• Data preview: Yes
• Column statistics: No
• Data owner: Yes
• Top data users: ?
• Change notifications: No
• Change feed: No
• Deployment:
• Supported data sources: S3, Kafka

Atlas

Website | GitHub

Apache Atlas is an open-source data catalog for metadata collection, governance, and data democratization.

More features

• Strategy: Push
• UX personalization: No
• AI autowiring: No
• Rich data profiling: No
• Recommendations: No
• Schemas, Description: Yes
• Complex schemas: No
• Data preview: No
• Column statistics: No
• Data owner: No
• Top data users: ?
• Change notifications: Yes
• Change feed: No
• Deployment:
• Supported data sources:HBase, Hive, Sqoop, Kafka, Storm

CKAN

Website | GitHub

CKAN is an open-source data catalog for data management, widely adopted by governments, NGOs, research institutions, and enterprises. It is actively maintained by a global community, with four of the eight core maintainers currently funded by Link Digital.

More features

• Strategy: Push, CKAN is used to publish or upload datasets into a central catalog
• UX personalization: Achievable through integration with CMS platforms like Drupal or any custom front ends.
• AI autowiring: Yes. Can be experimented using ckanext-embeddings.
• Rich data profiling: Yes, via ckanext-validation extension.
• Recommendations: Yes. Custom development or integration would be required.
• Metadata Schemas and Custom Fields: Yes, using ckanext-scheming​. One of the most widely used extensions for complex metadata schemas.
• Complex schemas: Yes (via ckanext-dcat extension)
• Data preview: Yes
• Visual previews: Yes
• Column statistics: No
• Data owner: No
• Top data users: Achievable using the stats-extension, which provides insights into user activity and dataset popularity.
• Change notifications: Achievable using ckanext-email-notification.
• Change feed: Yes
• Deployment: Yes (Self-hosted, cloud-hosted)
• Supported data sources: Yes; various formats:
  • Postgres
  • MySQL
  • SQL server
  • PowerBI
  • Tableau
  • CSV
  • Croissant
  • JSON
  • GeoJSON
  • XLS
  • Tableau
  • ...and more!

Magda

Website | GitHub

Magda is an open-source data catalog that features data discovery, metadata enrichment, and federation, focused on geodata.

More features

• Strategy: Push via UI
• UX personalization: No
• AI autowiring: No
• Rich data profiling: No
• Recommendations: No
• Schemas, Description: Yes
• Complex schemas: No
• Data preview: Yes
• Column statistics: No
• Data owner: Yes
• Top data users: ?
• Change notifications: No
• Change feed: No
• Deployment:
• Supported data sources: Mostly geodata

Open Data Discovery (ODD) Platform

Website | GitHub

First open-source data discovery and observability platform. ODD Platform is based on ODD Specification.

More features

• Strategy: Push/Pull
• UX personalization: No
• Rich data profiling: Yes
• Data collaboration: Yes
• Schemas, Description: Yes
• Complex schemas: Yes
• Data preview: No
• Column statistics: Yes
• Data owner: Yes
• Change notifications: Yes
• Change feed: Yes
• Metadata versioning: Yes
• SaaS: No
• Third-party integrations: Airflow, Apache Spark, Dbt, Great Expectations, and Prefect
• Supported data sources: Airflow, Athena, AzureSQL, BigQuery, Clickhouse, Databricks, DeltaLake, Druid, DynamoDB, Fivetran, Glue, Hive, Kafka, Looker, MariaDB, MlFlow, MSSQL, MySQL, Oracle, Postgres, Presto, Redash, Redpanda, Redshift, Snowflake, Tableau, and Vertica

OpenMetadata

Website | GitHub

OpenMetadata is the all-in-one platform for data collaboration, discovery, governance, lineage, and quality that lets you focus on building and analyzing.

More features

• Strategy: Push/Pull
• UX personalization: No
• Rich data profiling: Yes
• Data collaboration: Yes
• Schemas, Description: Yes
• Complex schemas: Yes
• Data preview: Yes
• Column statistics: Yes
• Data owner: Yes
• Change notifications: Yes
• Change feed: Yes
• Metadata versioning: Yes
• SaaS: Yes
• Third-party integrations: Dbt, Great Expectations, and Prefect
• Supported data sources: Airbyte, Airflow, Athena, AzureSQL, BigQuery, Clickhouse, Dagster, Databricks, DB2, DeltaLake, Druid, DynamoDB, Fivetran, Glue, Glue, Hive, Kafka, Looker, MariaDB, Metabase, MlFlow, Mode, MSSQL, MySQL, NiFi, Oracle, Postgres, PowerBI, Presto, Redash, Redpanda, Redshift, Salesforce, SingleStore, Snowflake, Superset, Tableau, Trino, and Vertica

Elementary

Website | GitHub

Apache Gravitino

Website | GitHub

Soda Core

Website | GitHub

Grai

Website | GitHub | Docs

More features

• Strategy: Push, Pull
• Customizable metadata model: Yes. The metadata model can be flexibly extended or modified as needed.
• Rich data profiling: No
• Recommendations: No
• Schemas, Description: Yes
• Complex schemas: Yes
• Data preview: No
• Column statistics: No
• Data owner: Yes
• Top data users: No
• CI Integration: Yes
• Lineage impact analysis: Yes
• Change notifications: Yes
• Change feed: Yes
• Automation: Yes
• UX personalization: Yes
• Deployment: docker-compose / Kubernetes with Helm, or using Grai SaaS offering
• Supported data sources:
  • Snowflake
  • BigQuery
  • Redshift
  • Postgres
  • MySQL
  • dbt
  • Slack
  • ... many others see the docs for a full list.

Hamilton

GitHub

A popular open-source framework for describing transformations that comes with a catalog (the Hamilton UI). The project originated from Stitch Fix. Created by Hamilton maintainers, Dagworks Inc. (YCW23) provides a managed version of the Hamilton UI in addition to self-hosted on-premise features.

More features

• Strategy: Use python to describe DAG, add one-line of code to capture: lineage & provenance, metadata, data summary profiles, versions, and execution telemetry. [See blog for details](https://blog.dagworks.io/p/hamilton-ui-streamlining-metadata).
• UX personalization: No
• AI autowiring: No
• Rich data profiling: Pluggable
• Recommendations: No
• Schemas, Description: Yes
• Complex schemas: Yes
• Data preview: Only summary statistics
• Column statistics: Yes, only for python.
• Data owner: Via tags.
• Top data users: Not yet
• Change notifications: Not yet
• Change feed: No yet
• Deployment: pip installable locally, self-hosted via docker, managed SaaS
• Supported data sources: No direct ingrations as of yet.

Egeria

Website | GitHub

📕 Proprietary Data & AI Governance Tools and Platforms

Collibra

Website | GitHub

Collibra is an enterprise data catalog that helps to discover and understand data that matters and drive impactful insights from it.

More features

• Strategy: Push
• UX personalization: Yes
• AI autowiring: ?
• Network-based: No
• Rich data profiling: ?
• Supported data sources:

Informatica

Website | GitHub

Informatica is an enterprise data catalog that provides AI-powered data discovery engine to scan and catalog data assets.

More features

• Strategy: Push
• UX personalization: ?
• AI autowiring: ?
• Network-based: Yes
• Rich data profiling: Yes
• Supported data sources:

Alation

Website | GitHub

Alation is a collaborative data catalog that helps companies to drive value and business impact from their data.

More features

• Strategy: Push
• UX personalization: Yes
• AI autowiring: No
• Network-based: No
• Rich data profiling: No
• Supported data sources:

Atlan

Website | GitHub

Atlan is a modern data catalog offering data discovery, data profiling, data quality, data lineage and data governance.

More features

• Strategy: Pull
• UX personalization: ?
• AI autowiring: ?
• Network-based: No
• Rich data profiling: ?
• Supported data sources: Presto, Deequ, Atlas, Airflow, Hudi

DataGalaxy

Website | GitHub

DataGalaxy is a modern data catalog offering data discovery, data profiling, data quality, data lineage and data governance.

More features

• Strategy: Pull & Push
• UX personalization: Yes
• AI autowiring: Yes
• Network-based: Yes
• Rich data profiling: Yes
• Supported data sources: [Available connectors](https://www.datagalaxy.com/fr/integrations-connecteurs/)

Talend

Website | GitHub

Talend is a data catalog that helps enterprises power critical business descisions with trusted data.

More features

• Strategy: Push
• UX personalization: Yes
• AI autowiring: ?
• Network-based: ?
• Rich data profiling: Yes
• Supported data sources:

Select Star

Website

Select Star is an intelligent data discovery platform that automatically analyzes and documents your data. Select Star provides an easy to use data portal that everyone can use to find and understand data.

More features

• Strategy: Pull
• AI autowiring: Yes
• Network-based: Yes
• Rich data profiling: No
• ER Diagram generation: Yes
• Role & Policy based access control: Yes
• Popularity & usage: Yes
• Description & Tag propagation: Yes
• Data preview: Yes
• Data owners: Yes
• Top data users: Yes
• UX personalization: No
• Supported data sources:
  • Snowflake
  • BigQuery
  • Redshift
  • Postgres
  • Looker
  • PowerBI
  • Tableau
  • Mode
  • Sigma
  • Sisense
  • Metabase
  • Looker Studio
  • DBT Cloud & Core
  • Slack

DataKitchen DataOps Observability

Website

More features

• Strategy: Full featured, with UI for singlue user. Enterprise version for teams
• UX personalization: No
• AI autowiring: DataOps TestGen data quality verification tool that does five main tasks: (1) data profiling, (2) new dataset screening and hygiene review, (3) AI/algorithmic generation of data quality validation tests, (4) ongoing production testing of new data refreshes and (5) continuous periodic monitoring of datasets for anomalies
• Network-based: Data Journey based
• Rich data profiling: 51 characteristics, with UI
• Supported data sources: Snowflake, Redshift, Tableau, Synapse, Postgres, Tableau, PowerBI, Airflow, Fivetran, Databricks, dbt, Databricks Azure Data Factory, SSIS, Synapse Pipelines, ADF-managed Airflow, Google Composer, AWS S3, Qlik Sense, Amazon Managed Workflows for Apache Airflow, Talend Cloud, Azure Functions (via Event Hub), Azure ADLS/Blob Storage (via Event Hub)

Monte Carlo

Website

Monte Carlo is a data observability tool that helps to increase trust in data by eliminating or preventing data downtime.

More features

• Strategy: Pull
• UX personalization: ?
• AI autowiring: ?
• Network-based: ?
• Rich data profiling: ?
• Supported data sources: Snowflake, Hive, Kafka, Looker, Redshift, Tableau, Big Query, Airflow, Fivetran, Presto, Mode, Periscope, Databricks, Glue, dbt, Chartio, Spark, AWS, S3, data.world, Google Cloud Platform

Databand

Website | GitHub

Databand is an observability platform that helps data engineers identify and troubleshoot pipeline issues and data quality problems.

More features

• Strategy: Push
• UX personalization: ?
• AI autowiring: ?
• Network-based: ?
• Rich data profiling: ?
• Supported data sources:

Datafold

Website | GitHub

Datafold is a data monitoring and observability platform that gives you confidence in your data quality through diffs, profiling, and anomaly detection.

More features

• Strategy: Push
• UX personalization: ?
• AI autowiring: ?
• Network-based: ?
• Rich data profiling: ?
• Supported data sources:

Ataccama

Website | GitHub

Ataccama is an enterprise data catalog and observability tool featuring data profiling and data quality management, designed for data professionals.

More features

• Strategy: Pull
• UX personalization: Yes
• AI autowiring: No
• Network-based: No
• Rich data profiling: Yes
• Supported data sources:

Collate AI Platform

Website | GitHub

DataHub Cloud

Website | GitHub

IBM Data Observability by Databand

Website | GitHub

IBM watsonx.data intelligence

Website | GitHub

Precisely

Website | GitHub

erwin

Website | GitHub

OneTrust

Website | GitHub

data.world

Website | GitHub

Zeenea

Website | GitHub

Explorium

Website | GitHub

Metaplane

Website | GitHub

BigID

Website | GitHub

OvalEdge

Website | GitHub

📒 Cloud Data & AI Governance Tools and Platforms

Amazon DataZone

Website

Amazon DataZone is a data management service that makes it faster and easier for customers to catalog, discover, share, and govern data stored across AWS, on premises, and third-party sources.

Google Cloud Dataplex Universal Catalog

Website | GitHub

Google Cloud Dataplex Universal Catalog is integrated platform that combines data discovery, cataloging, governance, quality, and exploration into one cohesive service.

More features

• Strategy: Pull
• UX personalization: ?
• AI autowiring: ?
• Network-based: No
• Rich data profiling: No
• Supported data sources:

Microsoft Purview Unified Catalog

Website

Microsoft Purview Unified Catalog is a central platform for discovering, classifying, and managing data assets across organization.

More features

• Strategy: Pull
• UX personalization: ?
• AI autowiring: ?
• Network-based: ?
• Rich data profiling: ?
• Supported data sources:

Databricks Unity Catalog

Website

Snowflake Horizon Catalog

Website

🔍 Data Observability Platforms

Back to top