Approach: Start with core requirements (fault tolerance, idempotency, eventual consistency), then design layered architecture with distributed tracing and append-only event stores for audit logging.

Considerations: Balance latency against compliance auditing; trade-off real-time consistency vs throughput; ensure regional data residency for regulatory compliance.

Key Points: Discuss sub-100ms payment processing (millions daily), immutable audit logs, and how BofA handles consumer and wholesale volumes while maintaining FDIC audit requirements.

Approach: Establish user impact scope via monitoring dashboards and logs; correlate backend logs with client telemetry; use distributed tracing to isolate root cause; execute fix with automated rollback capability.

Considerations: Critical services require minimal MTTR; balance quick patching against thorough root cause analysis; decide roll back vs roll forward based on severity.

Key Points: Reference BofA's 24/7 uptime SLAs, demonstrate incident severity classifications, and explain coordination with compliance if regulatory systems impacted.

Approach: Build real-time rules engine with streaming ingestion; combine rule evaluation with ML scoring; store decisions with audit context; enable business team configuration without code deployment.

Considerations: False positives harm customer experience; false negatives expose fraud risk; balance sensitivity tuning against operational burden and customer friction.

Key Points: Explain sub-second latency requirements, handling millions daily, and compliance for suspicious activity reporting (SAR) under AML/KYC regulations.