My work connects product intent, clear responsibility, and evidence of what actually happens. That applies to the organizations I lead and the systems I build.
I start with a human need and the decisions required to meet it. For a learning product, that means useful practice and feedback. For a connected workshop, it means one coherent workflow. For privacy review, it means a clear path to a sound decision.
The product intent gives architecture, staffing, and process choices something concrete to serve.
Governance needs to define who can decide, when someone should escalate, and where human intervention matters. People need understandable paths through the work and responsibility for the outcome.
At Meta, I applied that thinking through review frameworks, operating plans, runbooks, and cross-functional forums. In connected products, it shapes permissions, integration boundaries, and failure behavior.
I use specialized AI agents across research, design, engineering, testing, and operations. I define product intent, decompose the work, establish constraints, and direct parallel investigation and implementation.
I reconcile competing recommendations, resolve tradeoffs, and decide what is ready to ship. Using agents across multiple products makes clear priorities and evidence standards especially important.
The same principle applies to AI inside a product: people should understand the system's recommendations, retain meaningful control, and know who remains accountable.
A connected system brings software into contact with physical behavior. The console project required protocol analysis and controlled hardware tests to understand why a PC woke immediately after suspend.
I set the experience and constraints, directed the investigation, and established what evidence would demonstrate that the fix preserved intentional wake behavior.
Interested in building a product, improving an operating model, or discussing a leadership role?
I'd welcome a conversation about what you're working on and where I could help.