Register or log in to access this video

New York • September 8 & 9, 2027
Loved LDX3 New York? Pre-sale tickets for 2027 are now available.
AI is getting more expensive. Learn what the best engineering organizations do to make sure the spend pays off.
In many companies, AI spend already equals 20% of fully loaded headcount cost, and it’s only going to grow. That money buys real speed for individual engineers. But faster coding doesn’t automatically make an organization better at delivering software, and most teams have no clear way to tell whether the extra cost is paying off.
In this talk, Roman Musatkin, Head of Product & Design at Swarmia, shares what the most effective engineering organizations are doing differently. They’re building the system that builds the product: raising how much you can safely hand to agents, changing how teams work, and keeping quality high as more code and more actors move through the codebase.
He’ll introduce Swarmia’s Build framework, a practical way to look at AI leverage, engineering effectiveness, and return on engineering together, so you can make sure you’re focusing on what matters instead of chasing vanity metrics.
Key takeaways
You’ll walk away with:
- A practical framework that contains three questions that matter: are we creating the conditions for AI to succeed, are we delivering value faster and sustainably, and what business value is all this engineering producing.
- A concrete way to measure whether your AI spend is justified, including cost per pull request and AI cost per engineer, so you can bring a real number to your next budgeting conversation.
- Three culture shifts you can try right away: prototype an end-to-end version before you prioritize, hand engineers open-ended problems instead of finished specs, and run small live sessions where product, design, and engineering build together instead of in sequence.
Promoted Partner Content