AI Creators
Claude
How Spotify runs agents across 20M+ lines of code, with Niklas Gustavsson
Jun 29, 2026
Key Points
- 1.Spotify has successfully integrated agents to manage over 20 million lines of code.
- 2.The shift towards AI tools has transformed workflow, reducing manual editing.
- 3.Spotify's experience with LLMs showcases their potential for solving coding problems effectively.
Summary
The Evolution of Coding Workflows
The integration of AI tools like LLMs at Spotify has drastically altered coding practices. Users have moved from heavy reliance on manual code editing to a more streamlined process where AI handles significant portions of coding needs.
Utilization of Large Language Models
Spotify began experimenting with LLMs early on, facing initial challenges but eventually achieving inspiring results. These models not only support code generation but also enhance understanding of existing code, facilitating better solutions to problems.
Managing Extensive Codebases
With a codebase exceeding 20 million lines, Spotify has found that LLMs perform exceptionally within their large monolithic repositories. This capability allows engineers to pull insights from the entire codebase to enhance their work.
Automation of Maintenance Tasks
Spotify recognized the excessive burden of code maintenance relative to the growth of their engineering team. This led to the development of systems that automate routine tasks, significantly improving workflow efficiency.
Worth watching for
This video is for software developers and engineering leaders interested in the integration of AI technologies in coding practices.