Think inside your AI world.
How do I connect BigQuery to ChatGPT?
The official BigQuery remote server answers ChatGPT with queries, metadata and resources — the warehouse as a conversational surface.
Direct answer: connect the official BigQuery remote MCP server to ChatGPT. In ChatGPT, add it from the connector settings (developer mode); the tool's own docs carry the current path. Authorise your project and ChatGPT runs queries, reads metadata and inspects resources — warehouse questions asked as questions.
Warehouse answers without the console
What the events schema holds, what the cohort query returns, what changed in the dataset — resolved conversationally against the project. The console remains for heavy work; the requests queue shrinks for everything else.
Sensitive tables need a paper trail
Worked case: your team ratified a PII-query rule — ad-hoc queries touching PII tables require a ticket reference, no exceptions, after an audit sampled the query history and found casual reads nobody could explain. BigQuery logs every query; requiring the ticket first is data-governance law you imposed above the logging.
Through Unl the query conversation arrives with the law present: the proposed join touches the users_pii table, no ticket cited — against your rule, the unexplainable-reads audit attached. The paper trail starts before the query, which is the only place it can.
BigQuery to ChatGPT puts the warehouse in conversation; through Unl the conversation answers to the data-governance laws you ratified — PII reads get their ticket before they get their answer.
Reads through Unl arrive with measured context — in the presence of the decisions you’ve already settled. The reach lane is live: one box, paste anything. If it speaks MCP, Unl can reach it. Readings arrive unprompted, the data beside the criterion; Unl is a courier, not a warehouse, and keeps only your keys and the frame.
Read further
Questions people ask
What does the official server give ChatGPT?
Queries, metadata and resources from your authorised project — Google's official remote MCP surface, client-agnostic.
Does Unl see my query results?
Unl measures reads against your rules; it holds your decisions, never your tables. Substrate content stays yours, and warehouse data stays in the warehouse.
What data rules ratify well?
Access disciplines (PII tickets), cost disciplines (partition rules), definition authority (which model owns a metric). Law above the logs, served at read time.
What if you cannot tell whether a column holds personal data?
Then treat it as though it does until somebody decides otherwise, and record who decided. The uncertain column is the one that causes the incident, and the decision about it is more reusable than the answer: the next person meets a labelled column instead of the same question.
What this is
Think inside your AI world — you stay in command
Save the thoughts, decisions and targets worth keeping, each with its reasoning, carried into every AI session the moment they matter. A new unit of exchange between you and your AI: the Settled Why with standing that travels. Unprompted.
Your whole AI world. What you decided at the epicentre. It plugs into Claude, Claude Code, ChatGPT and Cursor as an MCP connector — quick to connect, in a couple of steps.
MCP native·Human settled·Model agnostic·Your data
Measured Context
Connect Unl to bring the right information into the moment.
Your sources, read against the criteria you set.
The full product, open. Free at launch.