This talk is one example of my work with leadership teams: helping them make AI decisions that don't just help a single squad, but multiply value across product, engineering, and operations.
About This Talk
This session is a practical, example-driven look at how a small AI pilot grew into a cross-functional force multiplier. We cover what worked, what changed quickly, and a repeatable playbook for leaders making similar decisions under real-world constraints.
The goal: leave with a clear picture of how AI decisions can be made, governed, and scaled—so they multiply value a year from now, not just during the demo.
Key Takeaways
- AI pilots can become force multipliers. See how an experiment that started in one team became a capability that multiplied value across product, engineering, and operations.
- Decisions under constraints are the real test. Learn how AI decisions were made while navigating risk, compliance, existing systems, and skeptical stakeholders.
- A repeatable playbook for leaders. Walk away with a framework for running AI pilots that don't stall after the demo—built for governance, monitoring, and long-term value.
- Cross-functional alignment is the unlock. Understand how product, engineering, and operations can see the same AI decision in their own terms and align on one shared move.
Slides
Slides are not available for public distribution at this time.
The Playbook
Practical frameworks discussed in this talk that you can apply immediately:
1. The Cross-Functional Lens
When evaluating any AI decision, ask: "Can product, engineering, and operations all see this in their own terms?" If not, alignment will stall and the pilot won't scale.
2. The Durability Check
Before committing to an AI capability, ask: "Can this be governed, monitored, and explained a year from now?" If the answer is no, you're building a stunt, not a capability.
3. The Constraint Gradient
Not all decisions carry equal risk. Map your AI decisions on a gradient from "safe to experiment" to "requires full governance" and allocate your resources accordingly.