No idea what's going on, can't report status to users and data in an inconsistent state? If you’re a software engineer that recognises these characteristics of a chaotic, un-orchestrated workflow then this is for you.

Based on a real production project, this live coding session adds order to chaos by introducing AWS Lambda Durable Functions as a resilient orchestration layer. I’ll walk through recording state across every part of the workflow, handling failure across different operations, debugging and versioning strategies.

We’ll add both unit and end-to-end tests (with LocalStack) to run in the CI pipeline and add instrumentation to show exactly what’s happening and where there are problems.

By the end of the session the workflow will be predictable, observable and reliable making the old chaotic version a thing of the past.