Serverless event-driven architectures promise automatic scaling and pay-per-use pricing, but delivering on that promise requires the right patterns and tools. In this session, we'll explore how Azure Container Apps and KEDA enable you to build scalable, cost-effective systems that scale from zero to hundreds of instances based on actual demand, whether you're processing queues, handling HTTP traffic, or running compute-intensive workloads.

We'll cover the fundamentals of how KEDA's event-driven scaling works, then dive into real patterns: queue-based processing, scheduled jobs, HTTP scaling, and combining multiple event sources. You'll see practical examples ranging from background job processing to GPU-accelerated workloads that only spin up when there's work to do. We'll also tackle the production reality: handling failures, managing cold starts, cost optimisation with scale-to-zero, and knowing when event-driven architecture isn't the right answer.

You'll leave with working knowledge of KEDA scalers and architectural patterns you can implement immediately.