Your business data deserves a clear explanation. Here is how the product is designed, and what we review with you before connecting your systems.
The query path validates SQL before execution and authorises access to the selected sources. Source database accounts should be configured with read-only permissions. Deployment credentials and network access are reviewed during onboarding.
Tenant identities are restricted to their assigned data sources. Platform administration is a separate capability. Current role controls govern application capabilities; do not assume per-role column masking or custom data restrictions without a separate review.
The product uses configured AI providers to interpret questions and build answers. Depending on configuration, prompts can contain business context and data. Provider processing, retention and hosting arrangements must be confirmed for your deployment.
Answers can include source scope and reporting context. The engine checks results and can ask for clarification. AI output should be reviewed for consequential decisions; the product does not promise perfect accuracy.
This page describes product design, not an independent security certification or a contractual service-level agreement. Deployment-specific commitments belong in your agreed service documentation.
