GenPage: Towards End-to-End Generative Homepage Construction at Netflix

Summary

Netflix traditionally constructs its structured two-dimensional homepage through a complex, multi-stage recommender pipeline that separates candidate generation and ranking across rows and entities. To simplify this architecture and optimize directly for whole-page user satisfaction, Netflix developed GenPage, an end-to-end generative transformer model that autoregressively builds the entire homepage from raw tokenized context. The system relies on a domain-specific tokenizer to compress engagement history, context injection for cold start, hybrid row decoding to minimize decoding steps, and reinforcement learning post-training. In online A/B testing against the production baseline, GenPage delivered statistically significant gains in core user engagement metrics while cutting end-to-end serving latency by 20 percent. Offline evaluations further showed that enriching context prompts improved recommendation quality more effectively than increasing model capacity in the current operating regime.

Context

Constructing the Netflix homepage is complex because it is a structured, two-dimensional layout of rows and recommendable entities where each choice affects others. Traditionally, Netflix built this page via multi-stage pipelines with separate candidate generation and ranking models at both the row and entity levels. This multi-model architecture introduced high operational maintenance, misaligned objectives across stages, heavy feature engineering, and challenges in optimizing for whole-page user satisfaction.

Approach / What changed

Netflix designed GenPage, a generative transformer that treats user engagement history and request context as a prompt to autoregressively generate the entire homepage layout, rows, and entities. GenPage employs domain-specific tokenization to compress sequences into discrete tokens, semantic embedding fusion and context injection for entity cold starts, constrained decoding for business rules, hybrid row decoding for inference efficiency, and pretraining followed by reinforcement learning post-training for page-level reward optimization.

Takeaways

• In online A/B testing against Netflix's multi-stage production recommender, GenPage delivered statistically significant gains in core user engagement while reducing end-to-end serving latency by 20%.
• Domain-specific tokenization compressed event representations—such as reducing a 16-token text sequence to 4 discrete tokens—lowering inference latency while retaining product control for business rules.
• Offline findings showed that enriching prompt context provided larger quality gains than increasing model capacity, and RL post-training increased page diversity without an explicit diversity objective.

Related reading

• Netflix · Jul 30, 2026

  GenRec: Towards LLM-Native Recommendation at Netflix

  Netflix developed GenRec, an LLM-backed recommendation ranker built by adapting an internal foundation model for large-scale personalization. Traditional recommendation stacks rely on thousands of hand-crafted features and specialized architectures that are costly to maintain across diverse content types. To replace manual feature pipelines, GenRec verbalizes user histories, metadata, and contexts into natural-language prompts and trains with multi-objective losses, including catalog-aware ranking and reward-weighted alignment. At inference time, the model executes in prefill-only mode on vLLM without decoding text. In large-scale online A/B testing against a mature production ranker, GenRec achieved statistically significant improvements in short-term and long-term metrics while using fewer labeled examples.

  Netflix Technology Blog

• Grab · Jul 3, 2019

  Making Grab’s Everyday App Super

  Grab manages an expanding superapp ecosystem comprising ride-hailing, food delivery, payments, and partner content surfaced through the Grab Feed. As content volume grows, the platform risks overwhelming users with irrelevant information. To address this, Grab built a recommendation engine that ranks cards using signals across user profiles, content metadata, and contextual factors such as time and location. The system employs multiple recommendation strategies—including popularity metrics, user favorites, collaborative filtering, habitual patterns, and cross-platform deep embeddings—which are selected or aggregated. Recommendation quality is evaluated via offline metrics like Recall@K and NDCG alongside online engagement experiments.

  Justin Bolilia

• Netflix · Jul 17, 2026

  In-House LLM Serving at Netflix

  Netflix established an in-house serving platform to run large language model inference directly inside existing production environments alongside traditional machine learning models. Built upon NVIDIA Triton Inference Server and vLLM, the unified architecture handles member-scale routing, candidate generation, feature fetching, and model execution over gRPC and OpenAI-compatible HTTP endpoints. The team shifted their primary engine from TensorRT-LLM to vLLM to support custom architectures, simpler debugging, and non-trivial constraint logic. Addressing production obstacles required patching Triton's frontend for guided decoding, pinning dependent library versions, rewriting logits processors in C++ for vLLM V1, and handling state machine resets during engine preemptions. The resulting platform unifies deployment flows while preserving operational stability across evolving model schemas.

  Netflix Technology Blog

• Meta · Aug 5, 2026

  From User Sequences to Scaling Laws: A Multi-Stage Architecture for Meta’s Ads Ranking

  Scaling sequence models for real-time ad recommendations faces latency bottlenecks and scaling ceilings when hybrid configurations process both long user histories and sparse feature interactions online. Meta addresses these trade-offs with a multi-stage architecture that decouples heavy, asynchronous offline user modeling from lightweight, real-time online ranking. Deep upstream transformer models process sequences of thousands of actions to generate cached user embeddings, which downstream online models combine with candidate ad features using dense tokenization and target-aware multi-head attention. This framework produces predictable, LLM-style log-linear scaling curves between compute and normalized entropy without disproportionate increases in online serving costs. Implemented as part of Meta's Generative Ads Recommendation Model, the design achieved a cumulative 6% lift in Instagram conversions and a 3% lift in Facebook conversions.

  By Steven De Gryze, Parshva Doshi, Sean O'Byrne, Arnold Overwijk, Dinesh Ramasamy, Lee Xiong