How would you design a highly available and fault-tolerant architecture for a cloud-based application?

Employed multi-region deployment, load balancing, and auto-scaling groups. Used services like Amazon RDS with read replicas for database resilience. Implemented health checks and graceful degradation to handle faults seamlessly.