Reviewing a full cycle for trends like team mood, talent signals, or biased language currently requires manual analysis across every individual review. This adds AI-generated insights at the cycle level, including improvement areas, talent signals, and bias-language checks, plus the ability for admins to ask free-form questions and get evidence-backed answers. Gives admins a faster, more thorough read on how a cycle went.