New framework makes governance conflicts explicit in multi-stakeholder software
Researchers propose a governance-architecture correspondence to prevent hidden bias in platforms.
Researchers Michael Nwankwo and Eric Umuhoza have proposed a new framework that bridges the gap between governance principles and software architecture for multi-stakeholder platforms (MSPs). Their work, published on arXiv, identifies that current software engineering patterns—such as data isolation and access control—ignore how conflicting stakeholder requirements should be prioritized when building platforms. Meanwhile, governance literature treats technology as neutral infrastructure, failing to guide architectural choices. The authors' governance-architecture correspondence framework explicitly maps five MSP governance principles to specific architectural decision spaces, such as visibility of data, decomposition of services, and algorithm design. For each principle, it defines the governance-aware design choice and the technically convenient default it overrides.
The framework is illustrated with a constructed knowledge platform for pig farming in Rwanda, where five stakeholder types (e.g., farmers, veterinarians, cooperatives, government agencies, and input suppliers) have structurally conflicting requirements—for example, farmers want private disease data, while authorities want transparency for disease monitoring. The framework makes these conflicts explicit and debatable before code is written. As work in progress, it has not yet been empirically validated; the authors plan a pre/post judgment study with platform users across all stakeholder types to test falsifiable predictions about governance outcomes. The paper is available on arXiv with the identifier 2605.31316.
- Framework maps five governance principles to specific architectural decisions (data visibility, service decomposition, algorithm design).
- Uses a pig farming platform in Rwanda with five conflicting stakeholder types as illustrative example.
- Not yet validated; planned pre/post judgment study with real users will test governance outcomes.
Why It Matters
Makes hidden governance biases explicit in software design, crucial for fair multi-stakeholder platforms.