How to Configure JetBrains AI Bigquery Connect

Published
Length
3 min read

Luis Felipe Dias

All posts

To set up jetbrains ai bigquery connect, link your cloud data source in the IDE. Then let the coding assistant read your active schema and run queries. Teams often use a shared service key or broad user role for fast setup. Broad roles let the assistant and the user read every table in the project. If a prompt triggers a bad query, it can scan huge tables or expose private rows. How jetbrains ai bigquery connect works in your IDE The IDE uses a local JDBC driver to talk to your cloud data war

Your coding agents run with your access.

Claude Code reaches prod with a developer's credentials, and approval prompts get clicked through. hoop.dev checks every query, masks sensitive data, and holds risky changes for a human.

To set up jetbrains ai bigquery connect, link your cloud data source in the IDE. Then let the coding assistant read your active schema and run queries. Teams often use a shared service key or broad user role for fast setup. Broad roles let the assistant and the user read every table in the project. If a prompt triggers a bad query, it can scan huge tables or expose private rows.

How jetbrains ai bigquery connect works in your IDE

The IDE uses a local JDBC driver to talk to your cloud data warehouse. You log in with a Google user account or a JSON service key file. Once you log in, the IDE pulls project, dataset, and table names into memory. The AI assistant reads this local schema map to help write SQL queries. When you ask the assistant a question, it sends table names to the model. You can also run the generated SQL right from the editor window. Every query runs with the exact cloud permissions of your logged-in identity. The cloud platform checks your role when the connection starts and when queries run. It does not check whether a human or an AI wrote the query.

Where native cloud roles fall short for AI queries

Cloud IAM roles are built to control which datasets a user can open. You can grant read roles on one dataset and block another dataset. This setup decides who logs in, not what individual statements do in flight. A developer often needs production read access to fix a live bug. When that developer turns on the AI assistant, the tool gets that same access. A single prompt can ask the IDE to pull sample rows from a table. Those rows might hold emails, phone numbers, card numbers, or national IDs. Native cloud roles return the full text unless you build complex policy tags first. Also, a wide select query can scan terabytes of data and spike your bill. Standard IAM roles cannot pause a risky drop or delete statement for peer review. Teams also struggle to see which queries came from AI prompts in cloud logs. Everything in the audit log just shows the developer's email or service key.

Steps to tighten native database access

Start by removing project-wide viewer and editor roles from daily developer accounts. Create a custom cloud role that only allows metadata reads and safe query jobs. Bind that role only to the specific datasets the developer needs this week. Replace downloaded JSON service keys with short-lived user logins from your identity tool. JSON keys sit on laptops forever and are easy for local apps to read. Turn on maximum bytes billed limits in your driver settings to stop huge scans. Finally, build authorized views that leave out sensitive columns before sharing tables with developers. These native steps reduce risk, though they take constant work to maintain across teams.

BigQuery IAM + AI Agent Security: architecture patterns and best practices

Continue reading? Get the full guide.

Free. No spam. Unsubscribe anytime.

Add runtime guardrails and masking beside your databases and APIs

Login rules, role grants, and CI checks govern who connects to your stack. Yet any allowed script, engineer, or AI agent still runs commands with the full scope of its role. For workloads on Postgres, MySQL, SQL Server, MongoDB, ClickHouse, Kubernetes, SSH, or internal HTTP services, hoop.dev sits beside the service as an open-source proxy, a sidecar for your database or API.

Before a query runs, hoop.dev guardrails check the statement against your rules. An AI Analyzer you configure can score intent or hold risky calls for human review. Next, hoop.dev data masking rewrites sensitive fields in responses, and each verdict is logged per listener. Explore the MIT-licensed project on GitHub.

Frequently asked questions

Does the JetBrains AI assistant read my raw table data by default?

It reads table and column names first to help write SQL queries. Still, users can attach query results or run AI actions that read actual rows.

Should developers use service account keys for IDE database connections?

No, static JSON keys on laptops are risky because they do not expire. Use personal user logins with short-lived tokens so every query links to a person.

How can we stop AI-generated queries from scanning huge tables?

Set a byte limit in your connection settings and require partition filters on tables. You can also block unscoped select queries before they reach the warehouse.

The query runs. The emails stay hidden.