GenRec: An LLM-Backed Recommendation Ranker at Netflix
Netflix's GenRec leverages LLMs to enhance recommendation systems, achieving significant performance improvements with fewer training examples.
Netflix is reshaping recommendation systems with large language models (LLMs) that enable richer modeling of users, content, and context in natural language. GenRec, built on an in-house foundational LLM, follows a two-phase framework to adapt and enhance recommendation ranking through user histories and context engineering. Results from A/B testing show significant gains in performance with fewer training examples, indicating a shift in the recommendation paradigm.
This synthesis was produced from its source by AI; there is no human editor or manual review step. How we work