backend-engineer

Bounded contexts in modular monoliths organize your backend into vertical slices, each with independent domain, application, and infrastructure layers. Use a folder structure that separates contexts by domain (auth, tax, BI, production), define APIs per context using Hono routing, and communicate between contexts via services rather than direct coupling to maintain isolation and testability.

Yes. Hono provides HTTP routing for backend endpoints, PostgreSQL with Drizzle handles data persistence, and a modular monolith structure lets you organize these across bounded contexts. This stack supports independent domain layers, repository patterns, use-case implementation, and dependency injection for testable, maintainable code.

Apply bounded contexts to isolate domains vertically, then split each context into domain, application, and infrastructure layers. Domain handles business logic, application coordinates use cases and DI wiring, infrastructure manages persistence and external communication. This separation prevents coupling and makes testing and maintenance straightforward across auth, tax, BI, and production domains.

Use a DI container to wire services within each bounded context, keeping dependencies scoped to their domain layer. Structure repositories and use cases to accept injected dependencies, then test in isolation using Vitest. This pattern ensures each context remains independently testable and reduces tight coupling between domains.

Yes. Use service-based communication between bounded contexts—auth, tax, BI, and production domains interact through defined service boundaries rather than shared databases or direct layer access. This maintains context isolation, enables independent scaling, and follows the duplication-over-coupling principle.