APIFrontends

APIFrontends are a design approach in which a dedicated frontend layer exposes APIs to clients, serving as the public face of an ecosystem of backend services. This layer can present tailored API surfaces for different clients such as web apps, mobile apps, or partner integrations, and may include data shaping, request orchestration, and protocol translation to reduce client complexity.

Architecturally, APIFrontends commonly involve a combination of an API gateway or edge proxy, a backend-for-frontend (BFF) layer, and a service composition layer. The API gateway handles routing, security, and traffic control, while the BFF layer provides client-specific aggregations and orchestration across multiple microservices. A service mesh or orchestration component may be used to coordinate calls, error handling, and retries. Authentication, authorization, rate limiting, caching, and telemetry are typically integrated to enforce policy and observe usage.

Technologies associated with APIFrontends span REST, GraphQL, and gRPC, as well as OpenAPI specifications for client contracts. GraphQL can be used in the BFF to aggregate data from multiple services, while REST or gRPC surfaces may be tailored to client needs. Observability through metrics, logs, and distributed tracing, along with security measures such as OAuth 2.0, JWT, and mTLS, are common.

Use cases include delivering mobile-optimized responses, reducing over-fetching, integrating partner APIs, and enabling per-client or per-device optimization. Benefits include improved performance, security, and developer experience, along with decoupled evolution of backend services. Drawbacks can involve added architectural complexity, duplicated logic across multiple frontends, and the need for robust governance and versioning.

See also: API gateway, Backend-for-Frontend, GraphQL, API management.