Architecture that holds when the decision matters.

I architect enterprise AI for a living. At Thomson Reuters, I work on global systems architecture for enterprise generative AI and multi-agent orchestration across legal, tax, and financial products—an environment where a wrong answer is not a bad demo; it is a liability.

That constraint shapes my approach: a single model returning a single confident answer is the wrong shape for consequential decisions. Better systems make disagreement legible, preserve the evidence behind a recommendation, and keep a human owner at each high-stakes gate.

The work spans strategic architecture, agentic-system governance, transformation, integration, and operating design. The through-line is simple: technology earns trust when people can understand its limits, challenge its reasoning, and own the choice that follows.

I also contribute to AI policy and executive communities where those standards are being worked out in public. In judicial contexts especially, governance stops being theoretical when a fabricated citation or opaque recommendation can change an institution's decision.