TL;DR

During its Q4 earnings call, Spotify co-CEO Gustav Söderström said the company’s best developers “have not written a single line of code since December.” The company credits an internal system called Honk, built on Claude Code, for enabling engineers to fix bugs and ship features from their phones.

Code From the Commute

Spotify shipped more than 50 new features and changes throughout 2025, including AI-powered Prompted Playlists, Page Match for audiobooks, and About This Song. The company attributes much of this velocity to Honk, its internal development system that enables remote, real-time code deployment using generative AI.

“An engineer at Spotify on their morning commute from Slack on their cell phone can tell Claude to fix a bug or add a new feature to the iOS app,” Söderström told analysts. “Once Claude finishes that work, the engineer then gets a new version of the app, pushed to them on Slack on their phone, so that he can then merge it to production, all before they even arrive at the office.”

The company described this as “just the beginning” of AI-assisted development.

Building a Unique Dataset

Beyond coding productivity, Spotify highlighted a competitive advantage that other AI companies cannot easily replicate: its proprietary dataset around music preferences.

Söderström pointed out that music-related questions rarely have factual answers. What counts as workout music varies by region, culture, and individual taste — Americans lean toward hip-hop, many Scandinavians prefer heavy metal, and European preferences cluster around EDM.

“This is a dataset that we are building right now that no one else is really building. It does not exist at this scale,” he said.

On AI-generated music, Spotify is allowing artists and labels to indicate in track metadata how songs were made, while continuing to police the platform for spam.

Looking Forward

Spotify’s experience suggests that AI coding tools are already changing how large technology companies operate, shifting the developer role from writing code to directing AI agents. For UK businesses exploring similar tools, the question is no longer whether AI can write production code, but how quickly internal workflows can adapt.