Talk Overview
Your AI pilot worked. Leadership loved the demo. And then… nothing.
Nearly two-thirds of organizations are stuck in pilot purgatory — AI experiments that proved feasibility but never reached production. The problem isn’t the technology. It’s that nobody defined what “ready to scale” looks like before the pilot started.
The Inverted 3 Horizons Framework
This talk introduces the inverted 3 Horizons framework — not as a maturity model, but as a metrics migration framework. Each horizon demands a fundamentally different type of measurement:
- H3 (Feasibility): Does it work? — The pilot phase most organizations never escape
- H2 (Value): Is it worth it? — Where most stalled pilots should be measuring
- H1 (Scale): Can we operate it? — Production readiness and organizational fit
Most pilots stall because organizations keep celebrating H3 feasibility (“it works!”) when they should be measuring H2 value (“is it worth it?”).
Key Topics Covered
- Gate criteria and time boxes for each horizon transition
- The three traps that kill the pilot-to-production transition
- How to distinguish technical validation from organizational readiness
- Evaluating blast radius and dependencies before scaling
- Warning signs that your initiative is stuck in the wrong horizon
Learning Outcomes
By the end of this session, you will be able to:
- Apply the inverted 3 Horizons framework to assess where your current AI initiatives sit and what’s required to advance them
- Define clear gate criteria for each horizon — including time boxes, success metrics, and exit conditions — before starting an AI project
- Identify the three common traps that stall AI initiatives at each stage and recognize the warning signs in your own organization
- Distinguish between technical validation and organizational readiness — understanding why proving “it works” isn’t enough to reach production
- Evaluate blast radius and dependencies to anticipate which teams, workflows, and systems will be affected by an AI implementation before scaling
Who Should Attend
Engineering leaders, product managers, and technical architects who have shipped AI pilots but struggle to move them into production. If your organization has a graveyard of successful demos, this talk is for you.
