AI BriefingSpotify Engineering news
The latest AI and developer news from Spotify Engineering, with short summaries.
Visit website →Spotify PMs: AI Product Management Requires Specialization Across Research, Model, and Experience Layers
Three Spotify Principal Product Managers explain how their roles differ across the research, model, and experience layers of the AI stack.
Spotify and NVIDIA Collaborate to Optimize LLM Inference for Recommendations
The collaboration delivers up to 54% cost reduction for beam search workloads on Qwen3 1.7B models.
Spotify launches Spotify Technology brand to unify internal developer tools
The new brand consolidates Portal, Confidence, Xirp, and Spotify for Backstage under a single identity for external sale.
Spotify: AI Adoption Doubles Change Volume Without Quality Degradation; Bottleneck Shifts to 'Verification'
Spotify analyzed that while AI adoption more than doubled change volume without quality degradation, the bottleneck has shifted to the 'verification' stage.
Spotify: Bayesian A/B Testing Is Not a Single Method… Default Settings Are Numerically Equivalent to Frequentist Approaches
Spotify explains that guarantees in Bayesian A/B testing vary depending on configurations such as priors and stopping rules, and that adding a separate Bayesian mode to its current experimentation program does not yield benefits that outweigh the operational costs.
Spotify's Portal Reduces Claude Code Token Usage by 90%
By leveraging Spotify's Portal AiKA Modes to route repetitive I/O tasks to low-cost models, token usage for Claude Code was reduced by 90%.
Conditions for LLMs to Replace Humans in A/B Testing
Using LLM predictions as a proxy metric for A/B testing requires specific statistical assumptions and calibration procedures.
Indexing a Data Lake for Online Point Queries
Spotify introduces a technique for indexing a data lake to enable low-latency data access for online services.
Spotify Adopts Kong's AI Gateway to Implement Generative AI at Scale
Spotify has adopted Kong's AI Gateway to efficiently operate generative AI services at scale.
Encoding Domain Expert Knowledge: The Context Layer Behind Spotify's Data Assistant
Spotify built an AI data assistant based on a context layer that combines domain expert knowledge to deliver accurate insights across a massive dataset.
Coding Is No Longer a Constraint: Spotify Expands Developer Experience for Teams and Agents
Spotify is revolutionizing developer productivity through AI coding tools and its in-house development agent **Honk**.
Building a Natural Language Interface for the Spotify Ads API with a Claude Code Plugin
Spotify made the Spotify Ads API controllable via natural language using a Claude Code plugin.
Background Coding Agents: Accelerating Downstream Dataset Migration (Honk, Part 4)
Spotify automated the migration of 1,800 pipelines using Honk and Backstage.
Spotify x Anthropic Live: Talking Agentic Development
Spotify and Anthropic shared the practical realities and organizational changes of agentic development.
The Technology Behind 2025 Wrapped Highlights
Spotify pre-generated 1.4 billion reports based on listening data from 350 million people.
A Multi-Agent Architecture for Smarter Advertising
Spotify built a multi-agent system using ADK and Vertex AI to automate ad media planning.
How Spotify Ships Its App: A Look Inside (Part 2)
Spotify shortened its release process by replacing Jira-centric manual work with a Backstage dashboard and Robot automation.
Congratulating the 2025 Spotify FOSS Fund Winners
Spotify selected FFmpeg, MSW, and Xiph.Org Foundation as the winners of the 2025 FOSS Fund.