The Vas Channel

I’m an analytically trained professional with a Political Science background and hands-on AI experience, focused on AI model evaluation through the lens of trust, safety, and ethical reasoning. I recognize bias and harmful content, and I bring rigorous, impartial assessment practices to help teams understand where AI outputs go wrong.

I apply structured evaluation rubrics to rate AI responses on factual correctness, tone, safety, and relevance. I also identify and document edge cases where outputs are misleading, harmful, or biased—then provide written justifications and recommended corrections, including structured annotation and harm assessment. Beyond evaluation, I’ve supported fact-checking workflows by assessing reliability of market and financial content for AI safety annotation.