Graph Analytics in Snowflake: Beyond Traditional SQL with Neo4j

Join Olga Razvenskaia, Senior Software Engineer in the Neo4j Graph Data Science team, as she demonstrates how to unlock powerful graph insights directly from your data stored in the *Snowflake Data Cloud* using the *Neo4j Graph Analytics* application from the Snowflake Marketplace.

Traditional SQL can struggle with complex, interconnected data. This session shows you a seamless, zero-ETL approach to modeling your tabular data as a graph and running sophisticated algorithms—all within your Snowflake environment.

Speaker: Olga Razvenskaia
View Presentation: https://drive.google.com/file/d/1GmJi…;

*What You Will Learn in This Session:*
The Zero-ETL Advantage: Understand why running graph analytics directly on Snowflake data, without the performance and cost drag of extracting, transforming, and re-importing (ETL), is the superior approach.

Anatomy of an Algorithm Call: Learn the three logical steps executed within a single SQL query in Snowflake:
– Graph Projection: Defining how to construct a graph from your node and relationship tables.
– Algorithm Execution: Running advanced algorithms like PageRank on the projected graph.
– Writing Results Back: Storing the computed results (e.g., node scores, embeddings) into new Snowflake tables.
– Modeling Tabular Data as Graphs: Dive into the technical requirements for your Snowflake tables (like requiring node ID, source node ID, and target node ID columns) and how to use SQL views to satisfy the schema requirements without data duplication.
– Live Demo: PageRank on Flight Data: Watch a hands-on demonstration using an open flights dataset to calculate the most influential airports, with results immediately written back to Snowflake tables and visualized using the Neo4j visualization package.
– Future Integrations: Get an exciting preview of the Snowflake Cortex AI Agent integration for conversational graph analytics, which simplifies complex configurational JSONs and helps analyze algorithm results in a chat window.

This is a must-watch for data scientists, data engineers, and analysts looking to leverage the power of graph algorithms on their massive Snowflake datasets using familiar SQL tools.

*Resources:*
– Get Started with Aura – https://bit.ly/3LOLrjh
– Deployment Center – https://bit.ly/4jOelM3
– Ground AI Systems and Agents with Neo4j – https://bit.ly/4oVsnyb
– Try the Application: Search for “Neo4j Graph Analytics” on the Snowflake Marketplace (Includes a 30-day free trial) https://app.snowflake.com/marketplace/listing/GZTDZH40CN/neo4j-neo4j-graph-analytics
– Official Documentation: https://neo4j.com/docs/snowflake-graph-analytics/current
– GitHub Examples: https://github.com/neo4j-product-examples/snowflake-graph-analytics

0:00 Introduction and Speaker Welcome
0:57 Understanding Snowflake: The Data Cloud
4:38 Important Disclaimer
7:01 Anatomy of an Algorithm Call (The 3-Step Process)
8:37 Compute pool selection
9:53 Modeling Tabular Data as Graphs
11:47 Schema Requirements for Node and Relationship Tables
13:17 Table of differences
14:55 Overview of Available Graph Algorithms
15:24 Result Types Overview
16:55 Estimation API
18:05 Visualisation Capabilities
18:45 Live Demo
22:13 The Job Life Cycle and Architecture
23:45 Future Integrations: Snowflake Cortex AI Agent. Upd.: Now available for preview https://medium.com/snowflake/announcing-the-private-preview-of-graph-agents-for-snowflake-intelligence-fbce69464798
22:15 Resources & Quick Links

#nodes2025 #neo4j #graphdatabase #graphrag #knowledgegraph