PuppyGraph

Zero-ETL graph query engine that queries existing lakehouses, warehouses, and relational databases as a graph with openCypher and Gremlin.

MCP

About PuppyGraph

Zero-ETL graph query engine that queries existing lakehouses, warehouses, and relational databases as a graph with openCypher and Gremlin. Explore how PuppyGraph integrates with the agentic data stack ecosystem and supports autonomous data operations.

Key Features

  • Zero-ETL: query relational and lakehouse tables as a graph without copying data into a graph store
  • Supports openCypher (Bolt protocol) and Gremlin query languages
  • Connects to Apache Iceberg (including REST catalogs), Delta Lake, Hudi, Hive, and Unity Catalog
  • Connects to warehouses and databases such as Snowflake, Databricks, BigQuery, ClickHouse, PostgreSQL, and MySQL
  • Graph schema defined as a mapping over existing tables, so one dataset can serve SQL and graph workloads
  • Distributed, vectorized execution with separated compute and storage for multi-hop queries
  • Web UI for schema building, query console, and graph visualization
  • Deploys in minutes via Docker, with Developer and Enterprise editions

Agent Integration

External Links

PuppyGraph MCP Server

Official MCP server for running Cypher and Gremlin queries and inspecting graph schema from agents

AI Integrations

Docs covering the MCP server, a text-to-Cypher chatbot demo, and custom agent query tools

← Back to Graph Database