The engineering productivity gap: How elite AI teams are pulling away from the rest

The distance between engineering teams that have operationalized AI and the teams that haven't is no longer a rounding error. The gap is real, it's measurable, and it's getting wider every quarter. If your org hasn't yet turned AI adoption into delivered work, you're already falling behind the teams that have.

Adoption on its own doesn't close the gap. If you measure AI-generated code the way you measure human code, it can even look like a problem, with bigger PRs and slower reviews. What separates elite teams is leverage, the ability to turn AI activity into merged, shipped work. Capturing it means seeing where AI is working, where adoption is stalling, and where you can replicate what your strongest teams already do.

In this session, Yishai Beeri, Andrew Zigler and Ben Lloyd Pearson walk through a mid-year benchmark refresh built on 2.7 million pull requests from 253 engineering organizations. You'll see how far the AI-elite teams have pulled ahead, why high AI usage correlates with a 1.7 to 2.2x lift in PR merge rate (roughly 1.5x for a typical team), and where that leverage leaks back out of the pipeline. You'll leave with the measurement play the strongest teams use to close the gap on purpose instead of by luck.

Register to join live and get first access to the full mid-year benchmark report the moment it's released.