BigQuery + MCP + Agents 🤖! BigQuery is now a supported data source for the MCP Toolbox for Databases. This integration allows your AI Agents with direct, standardized access to your BigQuery data via Model Context Protocol (MCP). What this means... ✅ Agent access to BigQuery via MCP ✅ Define analytical tools easily using bigquery-sql config ✅ Consistent tooling for BigQuery, AlloyDB, CloudSQL, Spanner and others ✅ Integration with Agent Development Kit Learn more with the links below. Wu Jiaxun Sandeep Karmarkar Wenxin Du Averi Kitsch Kurtis Van Gent Hangfei Lin
Quickstart with BigQuery: https://googleapis.github.io/genai-toolbox/samples/bigquery/local_quickstart/
Vinay Narayana, Sandeep K. I think the combination of the Google Agent Development Kit + support for BigQuery as a tool through Model Context Protocol Server are worth a look if you haven't explored it already. It is the most promising option I've seen so far for enabling natural language access to data in BigQuery. I was imagining it being useful for you if you wanted to set up an agent that business users could come to to ask any of the top x most common queries against each data product. I mocked up an example agent yesterday against BQ and found it very intuitive and liked that ADK comes with a react frontend you can deploy as well. I'm going to work on publishing my sample in a short blog post so I'll share it with you all once I do.
Google Cloud | Developer Advocate | Data Engineer
4dMore about MCP Toolbox: https://googleapis.github.io/genai-toolbox/getting-started/introduction/