The data mesh concept shifts the responsibility for managing analytical data from central teams to domains and treats shared data as data products. However, in heterogeneous enterprise environments, technical interoperability at data product boundaries remains challenging because domains tend to expose data in different formats and through different access modes. This results in repeated provider-side adaptations, consumer-side conversion logic, and additional coordination effort.

This thesis examines whether the above problem can be addressed by introducing a transmission layer as an optional self-serve platform capability in a data mesh. Based on the foundational concepts of data mesh, interoperability, data types, and data consumption patterns, the thesis derives requirements for such a layer and develops a corresponding concept. The transmission layer operates at data product output interfaces and provides alternative representations of data where this creates business value. It does so mainly through format conversion and by supporting different consumption patterns.

To prove the practical feasibility, a proof-of-concept prototype was developed for tabular and relational data. The prototype uses an internal hub format and supports multiple source and target formats as well as multiple consumer-facing access options. The concept and prototype are evaluated against the requirements and this evaluation is complemented by an expert interview.

The results suggest that a transmission layer is a viable approach for improving technical interoperability in a data mesh while preserving domain autonomy. At the same time, the work indicates that a production-ready implementation would require further work regarding governance integration, contract awareness, lifecycle management, semantic fidelity, and scalability.