The Content Optimizer is a Bayesian optimization engine designed to systematically refine website snippets and page content to outrank competitors in AI grounding responses. It operates through iterative cycles where specialized LLM roles propose strategic hypotheses, rewrite content, and evaluate ranking positions against real search competitor sets. By employing Bayesian acquisition policies—such as Thompson sampling, UCB, and ε-greedy selection—combined with a structured Rank Factor taxonomy, the engine balances exploring novel content variations with exploiting proven, high-performing changes.