Decentralized Data Ownership and API-Centric Access: Implementing Data Mesh Principles with PostgreSQL and Graph Databases

The increasing demand for agile, scalable, and federated data architectures has accelerated the adoption of decentralized data management paradigms. First introduced by Zhamak Dehghani in 2019, Data Mesh promotes domain-driven ownership, self-serve data infrastructure, and data-as-a-product principles to address challenges in modern distributed data ecosystems. This paper presents a practical implementation of Data Mesh principles, leveraging PostgreSQL for ACID-compliant transactional workloads and Graph databases (e.g., Neo4j, ArangoDB, Amazon Neptune) for efficient modeling and querying of complex relationships across distributed data domains. We propose an API-centric access layer that empowers data product teams to manage their data autonomously, while maintaining enterprise-wide governance, security, and compliance. Furthermore, we establish a security-first approach to implementing decentralized data architectures, outlining robust access control mechanisms, identity management strategies, and governance models to facilitate a seamless transition toward a domain-oriented and API-driven Data Mesh framework.