gcp Google Cloud Blog ·

BigQuery Graph Adds Measures Support for Agentic Workloads (Preview)

blogaigcppreviewarchitectretailgcp-bigquerygcp-looker
feature announcement

BigQuery Graph introduces support for "measures" in preview, unifying governed metrics with relationship mapping to enhance autonomous, agentic AI workloads. This allows agents to reason precisely across complex data dependencies, addressing the "why" behind business performance and reducing inaccurate insights from raw tables. The update also provides native tools in BigQuery Studio for visual graph modeling and Conversational Analytics integration. Deep integration with Looker ensures consistent semantic layers and streamlines DevOps workflows for graph intelligence.

  • BigQuery Graph introduces measures support for agentic workloads
  • New visual graph modeler and Conversational Analytics integration in BigQuery Studio
  • Native Looker integration ensures unified semantics and DevOps workflows
Features (3)
  • BigQuery Graph introduces measures support for agentic workloads

    Now in preview, BigQuery Graph allows data modelers to define "measures" (like SUM or AVG) directly within Property Graph DDL. This unifies governed metrics with relationship mapping, enabling AI agents to reason across complex dependencies with precision and preventing inaccuracies when working with raw tables.

  • New visual graph modeler and Conversational Analytics integration in BigQuery Studio

    BigQuery Studio now features a no-code, drag-and-drop visual graph modeler to simplify building and editing property graphs. Conversational Analytics agents can also leverage these graphs to convert natural language questions into precise, relationship-aware GoogleSQL or ISO GQL queries, preventing model hallucinations.

  • Native Looker integration ensures unified semantics and DevOps workflows

    BigQuery Graphs can now be integrated natively with Looker as in-database analytic models, supporting both database-managed and Looker-managed models. This centralizes business metrics at the data layer, ensuring consistent core KPI calculations and enabling enterprise DevOps workflows through the Looker IDE and Git-based version control.

Read the original announcement →

https://cloud.google.com/blog/products/data-analytics/bigquery-graphs-with-measures-for-trusted-agentic-workloads/

Related releases