Qry is a deployment and delivery layer for trusted business models. Your team remains responsible for deciding what data to publish and for the lawful basis and permissions behind that data.
Managed refresh
In managed mode, Qry stores the warehouse connection metadata and uses the configured credential during controlled refreshes. Use a read-only identity limited to the schemas or views needed by the endpoint. Public requests read a verified snapshot and do not run SQL against the warehouse.
dbt Push
In dbt Push mode, your repository or CI job runs dbt and materializes the release artifact. Qry receives the artifact and serving metadata, but does not receive the source-warehouse credential. If the artifact contains personal or regulated data, Qry still processes that hosted data as part of providing the service.
Credentials and logs
Qry API keys authenticate downstream callers and are separate from warehouse credentials. Full API keys are shown once; request and refresh logs are redacted operational records. Do not place credentials or sensitive data in URLs, query parameters, or frontend code.
For contractual privacy terms, retention, deletion, and subprocessors, see the Privacy Notice and contact us about a data processing addendum before sending production personal data.