Deploying Semantic ID-based Generative Retrieval for Large-Scale Podcast Discovery at Spotify

2026-03-18

AI Summary

Traditional recommender systems struggle to jointly model users’ long-term preferences and dynamic exploration intent. This work proposes GLIDE, a novel framework that formulates podcast recommendation as a semantic-ID-based instruction-following generation task. By discretizing the catalog into semantic IDs, leveraging large language models (LLMs) for semantic reasoning, and injecting user long-term embeddings via soft prompts, GLIDE enables efficient personalized generative retrieval. The approach effectively balances contextual awareness with low-latency inference, meeting industrial deployment constraints on cost and response time. In online A/B tests on Spotify, GLIDE increased engagement among non-habitual podcast listeners by 5.4% and boosted discovery of new shows by 14.3%, demonstrating its practical efficacy in real-world recommendation scenarios.

Technology Category

• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Natural Language Processing: Generation
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Personalized, context-aware and across-device search

Abstract

Podcast listening is often grounded in a set of favorite shows, while listener intent can evolve over time. This combination of stable preferences and changing intent motivates recommendation approaches that support both familiarity and exploration. Traditional recommender systems typically emphasize long-term interaction patterns, and are less explicitly designed to incorporate rich contextual signals or flexible, intent-aware discovery objectives. In this setting, models that can jointly reason over semantics, context, and user state offer a promising direction. Large Language Models (LLMs) provide strong semantic reasoning and contextual conditioning for discovery-oriented recommendation, but deploying them in production introduces challenges in catalog grounding, user-level personalization, and latency-critical serving. We address these challenges with GLIDE, a production-scale generative recommender for podcast discovery at Spotify. GLIDE formulates recommendation as an instruction-following task over a discretized catalog using Semantic IDs, enabling grounded generation over a large inventory. The model conditions on recent listening history and lightweight user context, while injecting long-term user embeddings as soft prompts to capture stable preferences under strict inference constraints. We evaluate GLIDE using offline retrieval metrics, human judgments, and LLM-based evaluation, and validate its impact through large-scale online A/B testing. Across experiments involving millions of users, GLIDE increases non-habitual podcast streaming on Spotify home surface by up to 5.4% and new-show discovery by up to 14.3%, while meeting production cost and latency constraints.

Problem

• generative retrieval
• podcast discovery
• semantic reasoning
• user intent
• large language models

Innovation

• Generative Retrieval
• Semantic ID
• Large Language Models
• Personalized Recommendation
• Instruction-Following