The world of CI/CD forced software engineers to be more disciplined and take ownership of testing.
Verifying that the code works as we expect, performs well and integrates with other systems is common practice but automatic check that acceptance criterias are met less so. Why is that? What is so hard about acceptance testing? How do we define it and where is its place within a CD pipeline? What can we learn from past mistakes and what approach works in practice? The talk will cover my experience with acceptance testing in two companies (LMAX and eBay), what pattern worked for us and how can we turn acceptance testing into a legitimate and effective tool.