PragAlign: Feedback-Guided Pragmatic Alignment for Controlled Synthetic Dialogue Generation

Synthetic dialogue generation can support research in privacy-restricted service settings, but generated conversations must preserve communicative intent, affective meaning, and natural dialogue flow. We introduce PragAlign, a feedback-guided framework for controlled synthetic dialogue generation conditioned on service context, target intent, and target emotion, with auxiliary trait-style controls. PragAlign uses a generate--evaluate--revise loop in which an LLM-based evaluator scores intent alignment, emotion alignment, coherence, fluency, and aggregate quality, then provides criterion-specific feedback for up to three refinement rounds. On 800 matched dialogue specifications, PragAlign achieves 99.50\% evaluator-defined acceptance, compared with 72.25\% for one-shot generation and 95.88\% for repeated generation without structured feedback.Repeated generation accounts for much of the improvement over one-shot generation, while structured feedback provides an additional gain in last-mile multi-constraint satisfaction. However, the largest apparent automatic improvement occurs in emotion alignment, where human judgments remain substantially less favorable. These results show that evaluator feedback can improve criterion satisfaction beyond repeated sampling, while also highlighting that same-judge gains may overstate human-perceived affective alignment.