Hey, I’m Derrick. I’m an AI Product Manager, Builder & Cross-border Strategist
I build AI workflows that turn ambiguity, pressure, and real-world constraints into structured systems people can actually use.
I’m Derrick, a builder working across AI product management, business development, ESG, and cross-border strategy.
My journey didn’t start with code. It started with messy commercial realities, stakeholder needs, and the challenge of turning ambiguity into something people can actually use.
AI Product Management · Agentic Workflows · ESG · Cross-border Strategy · Business Development · UofT MI ISD Incoming · CSC
I did not come into this work from a pure engineering path. My starting point was finance, business development, ESG, and cross-border commercialization — places where decisions rarely arrive cleanly and execution is often harder than strategy decks make it look.
I’ve seen how pressure, ambiguity, stakeholder misalignment, and operational friction can slow good people down. That experience shaped how I think about AI products.
I’m less interested in AI as a clever chatbot and more interested in AI as a system that helps people decide, move, and follow through.
That is a big part of why I built Alter AI and why I’ve been studying AI Product Management seriously: I want to build systems that help people handle uncertainty, reduce task-freeze, and navigate complex work with more clarity, structure, and momentum.
I don’t see AI Product Management as simply shipping intelligent features.
To me, AI PM is the work of giving a probabilistic system permission to affect the real world — while making sure it stays controllable, testable, recoverable, and useful.
My baseline assumption is simple: models fail, users hesitate, and edge cases dominate.
That is why I care about golden sets, red-teaming, graceful degradation, fallback behavior, and release-readiness thinking. Good AI products should not only sound intelligent. They should behave reliably under real constraints.
For me, the real question is not: “Can the model answer?”
It is: “Can the product still help when the model is uncertain, the user hesitates, or the situation becomes messy?”