AI Dev Live 2026: What to Watch and How to Measure AI’s Impact (Series #AI4SDLC)

Suppose an agent puts together a PR in ten minutes. Review and rework take another hour. Did we get faster?

On November 11–12 at AI Dev Live, we’ll talk about how to bring agents into a team’s work. If you’ve already tried an agent on a couple of tasks, here are three talks I’d start with.

🔸 Rodion Mostovoy — how to keep agents on task when the work stretches into hours. 🔸 Kirill Mokevnin — where to start in a legacy codebase and when to entrust an agent with more. 🔸 Alex Volochnev — what context an AI reviewer needs to catch breakage in a neighboring repository.

There are 18 online talks, starting at 14:00 Moscow time on both days. The full program is on the event site.

My talk, “AI Sped Up Development: How Do We Know It’s True?” — November 12 at 19:20 Moscow time I’ll look at why studies from Microsoft Research, METR, and Google get different answers and how to measure the effect in your own team. We’ll return to that PR and calculate how much engineering time it took, how fast the change reached users, and what the result cost after review, rework, and AI spending.

Come along! It’s free if you subscribe to the speakers’ Telegram channels. The organizers say they’ll share recordings and slides. 🔗 Registration and details

#AI4SDLC #Agents #Engineering #Metrics #Conference