Modern systems fail in ways that are often difficult to predict: dependencies slow down, networks become unreliable, services degrade under pressure, and assumptions that hold in normal conditions break at scale. In this talk, I will explore the principles of design for failure and show how to build systems that remain resilient when things go wrong.
Rather than focusing only on prevention, this session emphasizes preparation, containment, and graceful degradation.

We will discuss how to design architectures that assume partial outages, service instability, cascading failures, and operational uncertainty as normal conditions rather than rare exceptions. The talk will cover practical patterns such as redundancy, isolation, timeouts, retries with backoff, circuit breakers, fallback strategies, bulkheads, and observability as a design requirement.

Attendees will leave with a clearer understanding of how resilience is built into system design, not added after incidents happen. The session is especially relevant for engineers and architects responsible for distributed systems, cloud-native applications, and production environments where reliability is critical.