Data Mesh Architecture: Interoperability, Co-Operation, and Co-Regulation 

Abstract

This paper examines the emergence of data mesh architectures and data intermediaries as key components in emerging data economies. It explores how data mesh architectures enable the decomposition of monolithic data processing platforms into networks of specialized, interoperable data intermediaries. Data mesh architectures allow for more flexible, purpose-driven, operationally-distributed data processing. We explore the analogy between physical supply chains and "Data Supply Chains" to better understand the challenges and opportunities associated with integrating value flows across many distinct sites of governance authority, deriving insight into ways that the governance of data supply chains may be beneficially reconfigured. The paper argues that "Data Intermediary Networks" comprised of data products built on data mesh infrastructure can foster cooperation and co-regulation among data product operators, data providers and data consumers, leading to a dramatic increase in the variety of economically viable data products, along with a proliferation of new markets within which those products can be networked together into meshes with the capacity to generate new kinds of value. We outline the technical, operational, and governance architectures that underpin these networked data economies, highlighting how they can balance autonomy and accountability at multiple scales. By enabling a proliferation of specialized data products and services, data intermediary networks have the potential to unlock latent forms of value in data-and when such networks are allowed to emerge via "middle-out" cooperation and co-regulation, rather than top-down control, they may empower society to forge alternative data economies that are open, diverse and resilient.