Twenty-five years in financial services, turning hard, regulated-data problems into working software. Most of my time now goes to what large organizations need to actually use AI: the semantic layer, agent-ready interfaces, the operating model. I build the prototypes that prove it out.
Values any dropped item in Path of Exile 2 and tells you whether to sell, craft, or vendor it — live economy data, sourced crafting steps, and honest uncertainty labels on every recommendation. Built end to end: discovery, prototyping, and AI-assisted delivery.
The same golden-source field is named differently in every system, table, and feed, and scores of downstream consumers layer on their own taxonomies. I built the domain's first semantic layer — hundreds of fields across its data surfaces, enriched with governance context — because an agent can't reason over data an organization can't consistently name.
An end-to-end working prototype of a next-generation legal-entity system of record, built against the real requirements of an established incumbent system. Built from designer Figmas and a spec I derived myself, using custom AI agents constrained to an approved component system — production-shaped, not a throwaway.
A working data-quality application built solo with AI-assisted development, as a proof of what one product leader can now ship. Profiles millions of records against authoritative sources, scores and confidence-routes every issue, and puts an AI agent on the ambiguous cases — with a full analyst UI, dashboard, and audit trail.
FlowIt began with a focused product question: can a lightweight workspace connect product evidence, roadmap decisions, execution, publishing approvals, and results across a growing portfolio?
This is working software, not a finished platform. I am using it to discover which workflows are genuinely useful and will document how the product changes as those use cases emerge.
The question I get most: "where do I actually start?" So I wrote it down — providers, model tiers, tokens, apps vs. IDEs, why context beats clever prompts, and a real vibe-coded example you can follow along with, embedded live.
Open the guideAI is shortening the distance between an idea and working proof. That changes what product leaders can contribute, how teams work with engineering and UX, and what leaders need to understand firsthand.
Read the Field NoteI was born and raised in Charlottesville, Virginia, and started at Capital One at eighteen. I spent close to two decades there, growing from an entry-level job to Director and picking up a real passion for technology and technical execution along the way. I've kept investing in it since, through graduate work, certifications, and lately teaching myself to build.
I moved to Delaware about ten years ago, and today I'm a Vice President of Product Management at JPMorgan Chase, in reference data. The work I care most about now is how product management changes when AI is genuinely within reach: the foundations a large organization needs to use it, and the operating model that lets a team build instead of just writing specs. It's the part that doesn't depend on me being in the room.
Outside work, I've got four grown kids and a standing preference for being near the water. I fish, I'm outdoors whenever I can be, and I follow sports closely. Most of my best thinking happens away from a screen.