How Spotify runs agents across 20M+ lines of code, with Niklas Gustavsson

The integration of AI agents into software engineering workflows has fundamentally shifted development practices at Spotify, enabling a transition from manual maintenance to automated, agent-driven operations. By leveraging infrastructure like "Honk," engineers now automate complex tasks—such as library migrations and code changes—across massive monorepos containing over 20 million lines of code. This shift prioritizes robust test automation and verification loops, allowing for thousands of daily production deployments while maintaining high quality standards. Beyond core engineering, these tools have democratized prototyping, enabling non-engineers and executives to transform natural language ideas into functional mobile and backend prototypes within hours. This evolution demonstrates that significant productivity gains stem from foundational infrastructure investments rather than increased human effort, ultimately allowing teams to ship products faster and validate ideas with real-world data more efficiently.