Producing over 1,200 videos a month across 17+ languages, Nickelodeon Digital Studios has turned high-volume content into a testbed for scalable AI. In this keynote, John Pagano shares how his team built and deployed a custom AI editing workflow using vector-embedded search, retrieval-augmented generation (RAG), and a bespoke Adobe Premiere extension to transform raw footage into localized content at speed. You’ll learn: → How LLMs and vector search are used to surface the right moments from massive video libraries → What it takes to build a real-time, human-in-the-loop system for editing at scale → Lessons from integrating AI into a fast-moving, global production pipeline This is a rare behind-the-scenes look at LLM Ops in action, not just as a prototype, but as a key part of a production engine delivering over 20B organic watch minutes per month.