gcp Google Cloud Blog ·

Google details BigQuery Graph's GA and cross-cloud analytics for AI agents

blogaiawsgcpdatabrickssnowflakegapreviewdata-scientistfinanceretailgcp-bigquerygcp-spanner
announcement

Google's latest post details the general availability of BigQuery Graph, which brings native graph capabilities directly to BigQuery for unified analytics. It explains how this capability eliminates data silos and operational overhead, enabling petabyte-scale graph traversals using ISO-standard Graph Query Language alongside SQL. The article showcases enhancements like cross-cloud lakehouse integration with Databricks and Snowflake, and performance improvements for GQL queries. These developments support use cases from fraud detection to knowledge graphs for AI agents, impacting data teams across various industries.

  • BigQuery Graph for unified analytics and AI agent context
  • Broad adoption across industries for analytical and agentic workflows
  • Performance improvements and agentic ecosystem around BigQuery Graph
  • Cross-cloud lakehouse integration for unified data traversal
  • Enhanced GQL performance and query expressiveness
Notes (5)
  • BigQuery Graph for unified analytics and AI agent context

    This note describes BigQuery Graph, which integrates native graph capabilities into BigQuery to solve complex relational questions about connections. It supports petabyte-scale graph analytics with ISO-standard GQL and integrates with BigQuery ML/AI functions.

  • Broad adoption across industries for analytical and agentic workflows

    Data teams leverage BigQuery Graph for threat and fraud detection, supply chain digital twins, identity resolution, Customer 360, knowledge graphs for AI agent grounding, and network lineage/infrastructure management.

  • Performance improvements and agentic ecosystem around BigQuery Graph

    Since its preview, BigQuery Graph has seen improvements in its graph engine, making it faster and broader. An agentic ecosystem has been built, allowing AI agents to construct, query, and maintain auditable memory on graphs.

  • Cross-cloud lakehouse integration for unified data traversal

    The Borderless graph Lakehouse feature enables a single BigQuery Graph to span native BigQuery tables and open Iceberg tables in other clouds, including Databricks Unity Catalog and AWS Glue. This allows in-place traversal without data movement or ETL pipelines.

  • Enhanced GQL performance and query expressiveness

    BigQuery Graph's GQL now offers 2x faster execution since preview, with 100x faster undirected traversals and improved cycle detection. New CALL statements and extended subquery support enable more expressive queries and reusable functions.

Read the original announcement →

https://cloud.google.com/blog/products/data-analytics/bigquery-graph-connecting-data-and-ai-at-scale/

Related releases