Within this framework, the present research proposes the development and validation of a domain ontology applied to SAN, as a key component of the ecosystem architecture. The objective of the ontology will be to provide a logical and structured foundation that promotes the integration, querying, and reuse of data in the context of public policies in SAN. Based on this structured knowledge base, the use of a virtual assistant based on generative language models (LLMs) with a Retrieval-Augmented Generation (RAG) approach will be validated, capable of providing evidence-based recommendations and facilitating the monitoring of compliance with SAN targets. In this way, the aim is to strengthen the platform’s semantic layer and evaluate the impact of its integration on improving access to strategic information for public health decision-making.