Durable AI Agents — They're Failure Proof

🎥 Recorded live at the MLOps World | GenAI Summit 2025 — Austin, TX (October 9, 2025) Session Title: Durable AI Agents — They’re Failure Proof Speaker: Cornelia Davis, Developer Advocate, Temporal Abstract: When AI agents move into production, users expect one thing — reliability. But behind the scenes, distributed systems face constant challenges like network instability, rate-limited APIs, and unpredictable outages. In this live demo-driven session, Cornelia Davis from Temporal breaks down how to build durable AI agents that can recover gracefully from failure. She’ll walk through the core principles of fault-tolerant system design and demonstrate how Temporal’s workflow orchestration makes it possible to ensure agents complete tasks reliably, consistently, and transparently, no matter what goes wrong behind the scenes. Packed with practical demos, this talk reveals what it really takes to make your AI agents production-ready and failure-proof. What you’ll learn: • What makes AI agents fail in real-world production systems • Principles of durability and fault-tolerant design • How to recover from rate limits, network instability, and service failures • How to use Temporal to ensure end-to-end reliability • Why durable orchestration is the missing piece for robust agentic AI.