Multi-disciplinary creative director working at the intersection of symbolic systems, human perception, and generative AI. I develop interaction frameworks that stress-test LLMs beyond prompt-following—exploring when models reflect system constraints vs. when they exhibit emergent structural behavior. My practice spans from high-integrity creative outputs (lyrics, visuals, narrative arcs) to meta-level audits of GPT architecture, instruction misalignment, and symbolic recursion. I document these as case studies in model agency, constraint encoding, and session-context evolution.