20 years of assessing tax returns against IRS standards is functionally identical to evaluating AI-generated responses against quality rubrics — both require identifying errors, inconsistencies, and gaps in complex, nuanced outputs., Categorizing and labeling 100+ weekly specimen types in Cerner and Epic — each with specific identifiers, test codes, and handling instructions — is the same rule-based, precision labeling discipline used to tag images, video, and audio in AI datasets., Explaining complex tax law and phlebotomy procedures in plain language to diverse clients — from tech-savvy professionals to elderly patients — developed an expert sensitivity to clarity, tone, and helpfulness in communication, the exact dimensions used to rank AI responses., A BS in Computing Technology & Security provides foundational understanding of how data systems, algorithms, and software pipelines work — enabling faster onboarding into AI evaluation platforms and a more informed perspective when assessing AI-generated outputs., Five seasons as a remote Intuit contractor — managing caseloads independently, meeting quality benchmarks without direct supervision, and adapting to platform updates mid-season — directly mirrors the self-directed, task-based structure of AI evaluation freelance contracts.