Polyglot persistence lets microservices choose the best database for each workload, but today it is mostly static, manually configured, and hard to manage at scale. This paper presents GenAI-Augmented Polyglot Persistence (GAPP) — an architecture that uses generative AI and large language models (LLMs) inside the data orchestration layer. GAPP automates workload-to-database selection, supports natural-language query federation, adapts orchestration based on runtime patterns, and provides explainable data lineage. Our prototype shows 94% query accuracy, complete lineage coverage, and an 85% reduction in development effort compared to traditional pipelines. The results highlight how AI can enable self-governing data fabrics that unify DevOps and DataOps. We conclude with key research challenges and a roadmap for building AI-native database management systems.