How secure is iDBQuery?

iDBQuery's security story is grounded in architecture, the choices that make data exposure unlikely in the first place, rather than promises bolted on afterward.

• In-place querying. iDBQuery reads your sources to answer a question and builds one live model on the fly. There is no warehouse holding a duplicate of your data to secure separately.
• Data-stays-local deployments. Private VPC keeps everything inside your cloud account, air-gapped runs with no outbound connection at all, and the Desktop app keeps data on a single machine. The more sensitive the workload, the more locally you can run it.
• No training on your data. Your tables, documents, schemas, and questions are never used to train AI models or shared across customers.
• Your controls persist. iDBQuery uses your database credentials and respects your network boundaries; workspaces are isolated so only authorized users reach your sources.
• Auditability. Because every answer cites the exact source rows behind it, you can verify what the system did, which is its own form of security: no black-box numbers.

For a formal security review, we will walk your team through the data-flow, deployment, and access model in detail. Reach out to start that conversation.