Researchers propose governed agent architecture for natural-language dashboard generation over enterprise data

Researchers describe Text2Dashboard, a DataBrain-specific prototype that translates natural-language analytic requests into inspectable dashboards. The system combines schema-constrained model decisions with typed tools, persistent state, and deterministic Hooks for approval, audit, checkpointing, recovery, and failure handling. The pipeline resolves entities, discovers metadata, enforces read-only SQL, composes dashboards, and applies static checks, dynamic preflight, and browser inspection. Model inference proposes actions while deterministic software controls execution and records state transitions.

Evaluation on frozen real-DataBrain tasks and controlled Hook faults reports strict success rates of 6/8 on metadata and SQL tasks: metadata selection passed 4/4, all four SQL tasks met semantic criteria, and 2/4 met the exact output-column contract. The final release passed 4/4 single-panel dashboard tasks, one two-panel task, and one existing-dashboard refinement; a parameterised task exceeded its step limit. All ten fault scenarios met their specified outcomes without unapproved external side effects.

Authors note that model inference accounted for over 97% of observed runtime in every reported group. They emphasize that these small, DataBrain-specific results do not establish production readiness, general text-to-SQL accuracy, or an efficiency advantage over manual dashboard construction.