Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore how to build truly autonomous AI agents using event-driven architecture in this 15-minute conference talk that challenges the traditional request/response paradigm for LLM-powered systems. Learn about StreamNative Agent Engine, an innovative framework evolved from Pulsar Functions where every perception, plan, and action becomes an event, enabling more robust and scalable autonomous workflows. Discover how to implement declarative agent meshes that compile to run on event-driven functions, master workflow and agent orchestration using streaming joins and time windows, and handle back-pressure and retries without blocking LLM operations. Understand why event sourcing provides superior performance compared to synchronous REST APIs for autonomous workflows, explore essential patterns for implementing idempotency, observability, and safety rails in production environments, and gain valuable insights from real-world lessons learned while running agents at scale. Designed for AI engineers and platform teams integrating LLM agents with real-time data systems, this presentation provides practical strategies for building resilient, event-driven autonomous agent architectures.
