Question

A financial services company operates a customer-facing loan application platform. Architects are designing for resilience against partial failures of downstream dependencies (fraud scoring service, credit bureau API, and a document-storage backend). Leadership requires that a dependency outage never cause a total platform outage, and that core functions remain usable even when non-essential services are unavailable. Which design approach BEST meets these requirements?

Accepted Answer

Implement graceful degradation with circuit breakers, timeouts, and fallback logic so that when a non-essential dependency fails, the platform disables only the affected feature while core functions continue

Answer

Configure synchronous retries with exponential backoff on all dependency calls so failed requests are eventually completed before returning to the user

Answer

Deploy the entire platform in an active-active multi-region topology so a regional failure automatically fails over to the second region

Answer

Increase the connection pool and timeout thresholds for each dependency so slow responses do not immediately return errors to callers