“How do we deliver more with fewer people?” At Agile Cambridge, my audience voted this the most important question they face today. And the obvious answer seems to be AI.
But what if I told you that work got faster with AI, and delivery still isn’t enough?

I’ve seen teams fix a bug with AI in less than a day, only for the fix to wait weeks for an approval meeting. AI made the work faster, but it didn’t make delivery faster. The speed went somewhere, and in most teams, nobody knows where.

In this talk, I’ll connect recent research on where AI speeds up work, and where it doesn’t, with real situations from my own practice. To make it practical, I’ll introduce a technique I call the Waiting Map, which participants can use the very next day to identify where work gets stuck and slows down the whole system, even when individual tasks get faster. These are the places where AI can actually make a difference, instead of speeding up the steps that were already fast.

No tool demos. One real problem, one method, and one question you can take back to your team on Monday: not how long did it take, but how long did it wait?