Boundary Conditions for Continuous Integration Delivery Reform in Kenya: Process Tracing of Accountability and Implementation Failure

Kenya's public-sector digital transformation agenda has embraced continuous integration and delivery (CI/CD) as a technical solution to chronic delays in government service platforms, yet repeated reform initiatives have stalled or produced fragile systems that fail under routine use. the present analysis argues that the dominant implementation discourse misreads the problem as purely technical, obscuring the accountability relationships that determine whether CI/CD practices can take root in a public-administration context. Through a process-tracing analysis of the institutional, political and administrative conditions surrounding Kenyan e-government projects, the study identifies three boundary conditions that jointly constrain reform: fragmented authority over infrastructure procurement, performance incentives that reward project launch rather than operational continuity, and weak feedback loops between citizen-users and system maintainers. The analysis synthesises theoretical work on public-sector reform, information systems implementation and accountability politics to propose a framework for diagnosing why CI/CD adoption fails in similar administrative settings.