Architecture Decisions — Azure Governance Platform

Decision to use specialized AI agents for different platform concerns.

Restricting tool access based on agent role and responsibility.

Choosing bd (beads) for git-native, local-first issue tracking.

Requiring evidence-based research before architectural decisions.

Implementing configurable compliance rule engine.

Mapping governance controls to regulatory frameworks.

Evolution path from B2C to direct Azure AD integration.

Migration from ACR to GitHub Container Registry.

Azure SQL tier selection and scaling strategy.

Data synchronization reliability with ghost job cleanup, staggered scheduling, and failure alerting.

Role-based access control with 35 resource:action permissions and 4 predefined roles.