This page shows how I train judgment. Across formal study, field work, and applied systems, I focus on improving how decisions get made when information is incomplete and outcomes carry consequence.
I return to structured environments when the questions require more precision.
Early study established the pattern. I was drawn to environments that forced clarity, even before I knew where it would lead.
I enrolled my team at Yahoo! into a course focused on improvisation. The objective wasn’t performance. It was collaboration: learning how to build on each other’s thinking, articulate ideas in motion, and design without shutting down possibility.
Coursework through the Institute for Computational and Mathematical Engineering (ICME) marked a deliberate turn toward data and systems. I focused on how decisions get made when information is incomplete, and how reasoning breaks under pressure.
Sustained study through MIT programs has focused on design, AI, and decision systems. This work has been self-directed and ongoing, shaped through continuous study and ambient learning rather than formal milestones.
Field work translates real environments into usable patterns for designing trust, interpretation, and accountability.
Across financial services, healthcare, and enterprise infrastructure, the work clarified how semantic systems, lineage, and cognitive legibility help people trust complex data environments.
As automation moved from deterministic rules toward adaptive systems, the work focused on explainability, confidence calibration, and human oversight inside operational workflows.
Inside Wells Fargo AI Enterprise Solutions, the work focused on operationalizing trust in regulated AI systems through governance frameworks, human oversight, interpretability, and scalable experimentation across 17 business units.
Inside the CA Accelerator, design became infrastructure for experimentation, validation, and trust across emerging technology ventures working in NLP, predictive DevOps, and operational analytics.
Inside advanced semiconductor design tooling, the work focused on user research, in-canvas interaction, UX maturity, and early machine-learning-assisted workflows for engineers operating at the edge of technical complexity.
Public systems rarely fail at a single touchpoint. This field study looks at how navigability, dignity, and operational clarity shape trust when people encounter institutions under pressure.
For Skitch, we observed preschool children fingerpaint to understand how people explore color before constraint. The work shifted markup palettes from rigid selection toward fluid exploration.
Embedded agency work across Fidelity and other enterprise clients showed how mobile strategy, behavioral segmentation, and content targeting shaped trust as digital systems began following people everywhere.
Through Stanford improv training, we practiced building on each other’s ideas while designing Yahoo! frontdoor experiences at global scale. Even fractional A/B tests reached millions of users.
Inside the first real social network, unstable infrastructure, shifting product identity, global usage patterns, and early social graph behavior revealed how quickly users could define a platform faster than the company itself.
Work across personalization systems, messaging convergence, and large-scale onboarding helped transform the internet from a technical frontier into an environment ordinary people could navigate with confidence.
Inside the dot-com invention factory, design functioned as early-stage scaffolding for startups translating everyday life into the internet before the infrastructure, economics, and ethics had caught up.
Learning proves itself only when it changes how decisions are designed, challenged, and held accountable.
Separate signal from output
Applied in systems where outputs must be interpreted under real conditions.
A practical operating model for separating signal, interpretation, escalation, and response when AI systems produce outputs that require human judgment.
Design for human interpretation
Applied where human judgment is designed into systems with ambiguity, escalation, and accountability.
A civic systems argument for rebuilding public infrastructure around human agency, interpretation, and dignity under real-world constraints.
Design systems that stay accountable when wrong
Applied in systems that must remain legible, correctable, and trustworthy under pressure.
A reflection on the disciplines that make systems legible, durable, and accountable when interpretation carries consequence.
Learning only matters if it improves how decisions hold under real conditions.
Learning is not accumulation. It’s calibration. It’s how I train for decisions that don’t have clean answers.
Every environment above, classroom, product, or civic system, forced the same adjustment: less assumption, more observation, and systems that stay accountable when they’re wrong.
I see that same pattern every day in how my son learns—curiosity without hesitation, questions without constraint. It’s also why I’m intentional about where I am and who I’m around.
I want decisions that hold up when it actually matters. That’s what I’m learning to do.