DEJAN captures and parses raw grounding metadata across major LLM providers—such as Google search grounding chunks and OpenAI response annotations—to differentiate between selected citations actually referenced in generated text and unselected sources merely retrieved during web browsing. This citation intelligence is paired with Bayesian snippet optimization and veracity scoring to systematically test content variants, ensuring high-fidelity brand representation and maximizing the conversion of retrieved pages into cited sources.