Most teams already have plenty of tests. They have unit tests, integration tests, UI automation, and impressive code coverage numbers. Yet production issues still happen, and a green CI pipeline doesn't always mean the software is ready to ship.

The challenge isn't writing more tests, it's knowing which tests actually increase confidence.

In this session, we'll move beyond the traditional "unit vs. integration" debate and explore how modern engineering teams build confidence across the software delivery lifecycle. We'll discuss where different testing approaches provide the most value, why code coverage can be misleading, and how techniques like contract testing, observability, feature flags, and production validation complement traditional automated testing.

Using practical examples from real production systems, I'll share how teams can reduce flaky tests, shorten feedback cycles, and build delivery pipelines that provide meaningful confidence instead of simply reporting more passing tests.

Whether you're building desktop applications, cloud services, or game engines, you'll leave with a practical framework for choosing the right testing strategy for the problem you're trying to solve not just the one that's easiest to automate.