AI Engine
What sits between the question and the SQL
A schema says which columns exist. It does not say which of four date fields means what a person means by “ordered”. The engine is the layer that closes that gap, and shows its working at every stage.
Ask your data anything,get an answer in seconds
Connect a database, ask in plain English, and ChatPlotDB writes the SQL, runs it, and hands back clean tables and charts your whole team can read.
What would you like to know?
A config, not a prompt
Which tables matter, which joins are real, and what your team's words map to, held as configuration and read at planning time. Change a definition once and every future answer changes with it.
It asks rather than guesses
When the metric you named resolves to more than one thing, the planner comes back with a question. An ambiguous question answered confidently is the most expensive output this system can produce.
Read-only, then verified
Generated SQL can be shown before it runs and always executes read-only. A verification pass then checks the result for the silent failures — a bad cardinality, an inner join that dropped the rows you asked about, a filter that matched nothing.