BigQuery meets ADK & MCP: Accelerate agent development with BigQuery's new first-party toolset

AI agents interacting with enterprise data systems like Google Cloud BigQuery require robust tools. Previously, developers built custom, time-consuming, and risky tools for this integration. Google now offers first-party tools for BigQuery to simplify this process. These tools include functions to fetch metadata and execute SQL queries. Developers can use these tools within Google's Agent Development Kit (ADK) or the open-source MCP Toolbox for Databases. Using these pre-built, Google-maintained tools allows developers to focus on application innovation rather than infrastructure. The article provides a tutorial on building a conversational analytics agent using ADK and BigQuery's first-party tools. The agent queries BigQuery's public e-commerce dataset to answer natural language questions. The tutorial details setting up an ADK application and integrating the BigQuery toolset. An alternative approach involves using the MCP Toolbox for centralized tool management and security. The MCP Toolbox decouples agents from direct BigQuery interaction, simplifying updates and security policy enforcement. Both methods enable agents to securely and intelligently access and query BigQuery data.