Date & time
17:00
Register for the panel discussion
Login or join LeadDev.com to view this content
AI coding agents can generate code faster than engineering teams can validate it. As the volume and velocity of AI-generated code increase, traditional indicators of quality, like passing builds and successful test runs, are no longer enough. Engineering teams need greater visibility into whether their tests are actually exercising the code being introduced and whether critical functionality is being adequately validated.
For engineering leaders, the question is shifting from “How much faster are we coding with AI?” to “How confident are we in the quality of the code AI produces?”
Code coverage is becoming an increasingly important quality signal in this new era of software development. But coverage alone doesn’t guarantee quality. A high coverage percentage doesn’t necessarily mean the right scenarios were tested, and a passing test suite doesn’t always mean a change is safe to ship.
In this session, we’ll explore how code coverage fits into the broader quality feedback loop for AI-first engineering organizations. We’ll also discuss how engineering leaders can move beyond coverage as a single metric and turn it into an actionable quality signal embedded directly into CI workflows.
Key takeaways:
- Cut validation time by running, and measuring coverage on, only what a given change affects
- Set merge-blocking quality rules and risk-based review levels, including which decisions can be automated
- Feed check results back to coding agents so they fix failures before human review