What happens when iDBQuery can't answer a question?
If iDBQuery can't answer, it tells you why rather than fabricating a number, whether the data isn't connected, a column doesn't exist, or the question needs clarifying, and it suggests what to connect or rephrase. It never returns an uncited figure it can't trace to source rows.
When iDBQuery cannot answer, it says so and explains why, rather than inventing a plausible-looking number: - If the data is not connected, it tells you which source you would need to add. - If a column or table does not exist for what you asked, it says the data is not available rather than substituting the wrong field. - If the question is ambiguous or under-specified, it asks a clarifying question.
A worked example: you ask 'what's our churn rate?' but no subscription or cancellation data is connected. iDBQuery explains that it cannot compute churn without that source and suggests connecting it, instead of returning a fabricated figure.
The differentiator is honesty by design. Because every real answer is cited to source rows and shown with its SQL, iDBQuery has no way to pass off an uncited guess as a result; if it cannot trace a figure to your data, it will not present it as fact. That is the opposite of a general chatbot, which will happily produce a confident but ungrounded number. For genuinely hard-but-answerable questions, the Analyst agent will attempt a multi-step investigation before concluding, and if it still cannot get there, it reports what it tried and what is missing. So a 'no' from iDBQuery is trustworthy, and usually comes with the next step to get to 'yes.'
Updated 2026-06-22