From prototype to principle: A design science approach to artificial intelligence innovation in maternal health

This study applies the Design Science (DS) methodology to the development and evaluation of a conversational artificial intelligence (AI) platform to support pregnant and postpartum women's informational and emotional needs. Motivated by persistent gaps in personalized, trustworthy, and context-aware maternal care technologies, the study combines artifact creation with iterative user engagement to derive practical and theoretical contributions. Over six weeks, thirty participants interacted with the AI system, generating 1421 unique queries. Analysis of interaction logs, semi-structured interviews, and usability surveys revealed a consistent demand for affectively attuned and informationally accurate responses that provide emotional validation. This work presents emergent design principles including traceable content, empathetic tone, modular depth, contextual personalization, and emotional responsiveness. The study introduces six design principles aimed at building emotionally intelligent conversational AI for maternal health. Additionally, it illustrates how DS can yield design knowledge transferable to other emotionally sensitive care environments.