Resource Hub
Explore the open-source ecosystem powering the agentic data stack. Discover components, their agent integrations, and real-world patterns. Learn what makes a data component truly agent-ready in our 6-Principle Evaluation Framework.
Catalog Service
Metadata catalogs that provide unified governance, discovery, and interoperability across data lakehouse engines and formats.
Explore →Lake Format
Open table formats that bring ACID transactions, schema evolution, and time travel to data lakes.
Explore →SQL Engine
Distributed and embedded SQL engines that power analytics queries across lakehouse storage.
Explore →Semantic Layer
Metric and semantic layers that provide consistent business definitions for agents and humans alike.
Explore →ETL & ELT Tools
Data integration and transformation tools that move and shape data across the modern stack.
Explore →BI Tools
Open-source business intelligence platforms for data visualization, dashboards, and exploration.
Explore →Data Agents
Agent frameworks and implementations that bring autonomous capabilities to data engineering and analytics workflows.
Explore →Scheduler
Workflow orchestration and scheduling platforms that coordinate data pipelines and task execution at scale.
Explore →Graph Database
Graph databases and graph query engines that model relationships natively, powering GraphRAG, knowledge graphs, and multi-hop reasoning for AI agents.
Explore →Agent Integration Landscape
The modern data stack is evolving to support autonomous agents. Components across every layer are adding MCP servers, CLI interfaces, and agent-specific skills.
- 29
- MCP-Enabled Components
- 31
- CLI-Enabled Components
- 32
- Total Components Tracked
Common Questions About Agentic Data Stack Tools
Everything you need to know about building with open-source agentic components.
What tools are included in the agentic data stack directory?
We catalog 27+ open-source tools across 8 categories: Catalog Services (Unity Catalog, Apache Polaris, Gravitino), Lake Formats (Apache Iceberg, Delta Lake, Apache Hudi), SQL Engines (Trino, DuckDB, Apache Spark), Semantic Layers (Cube, dbt Semantic Layer), ETL/ELT Tools (Airbyte, dbt, Apache Flink), BI Tools (Apache Superset, Metabase), Schedulers (Apache Airflow, Dagster), and Data Agents (DatusAI, WrenAI).
How do I choose the right tools for my agentic data stack?
Use our 6-Principle Evaluation Framework to assess tools: (1) API-First Design - programmatic access for agents, (2) Declarative Configuration - YAML/JSON over imperative code, (3) Semantic Awareness - business logic understanding, (4) Versioning & Time-Travel - rollback capabilities, (5) Observable & Explainable - audit trails for agent actions, and (6) Self-Service Ready - minimal human intervention required.
Are these tools production-ready?
Yes, all tools in our directory are production-grade and battle-tested. Tools like Apache Iceberg, Trino, dbt, and Airbyte are used by thousands of companies including Netflix, Uber, Apple, and Airbnb. We focus on mature projects with active communities, not experimental frameworks.
Can I use these tools with cloud data warehouses like Snowflake or BigQuery?
Absolutely. Most tools support hybrid architectures. For example, Trino and DuckDB can query Snowflake/BigQuery directly, Apache Iceberg works with S3/GCS/ADLS, and dbt supports 30+ data platforms. You can mix cloud-native services with open-source components based on your needs.
How do I get started building an agentic data stack?
Start with three core components: (1) Choose a lake format (Apache Iceberg or Delta Lake) for versioned storage, (2) Set up a catalog service (Unity Catalog or Apache Polaris) for metadata management, and (3) Deploy a SQL engine (Trino or DuckDB) for flexible querying. Then layer on transformation tools (dbt), orchestration (Airflow), and data agents as your needs grow.