Spotify Engineers Unveil Revolutionary AI-Powered Ads Manager Built with Claude Plugins

Breaking: Spotify Launches Conversational Ads Manager Using Natural Language

Spotify Engineering has announced a groundbreaking tool that allows advertisers to manage campaigns via plain English commands, bypassing traditional code-heavy interfaces. The system, built using Anthropic's Claude Code Plugins, transforms OpenAPI specifications and Markdown documentation into a fully conversational ad management experience—without a single line of compiled code.

This development marks a major leap in ad tech accessibility, enabling non-technical marketers to interact with Spotify's Ads API as naturally as chatting with a colleague. Spotify engineers confirmed the tool is already in internal testing and could reshape how digital advertising campaigns are run.

“We wanted to see if we could turn a complex API into something you can just talk to,” said Sarah Chen, lead engineer on the project. “With Claude plugins, we fed it the OpenAPI spec and some Markdown docs, and it just worked. No compilation, no deployment headaches.”

Background: The Challenge of Ad Tech Complexity

For years, managing programmatic advertising on platforms like Spotify required deep technical knowledge—reading API documentation, writing code, and debugging integrations. Marketers often relied on developers to translate campaign goals into API calls, causing delays and miscommunication.

Spotify's Ads API serves over 500,000 advertisers globally, but its complexity created a barrier. Engineers at Spotify sought a solution that would let anyone describe their campaign in natural language (e.g., “Show my new album ad to hip-hop fans in New York this weekend”) and have the system automatically generate the correct API request.

Enter Claude Code Plugins, an extension of Anthropic's Claude AI that can interpret structured data and execute code. Spotify's team experimented with feeding the AI the full OpenAPI 3.0 specification of their Ads API, along with markdown-based usage guides.

“The key insight was that Claude could understand the API contract as a conceptual framework,” explained Dr. James Park, a UX researcher at Spotify. “It didn't need a custom wrapper—it just reasoned about the endpoints, parameters, and authentication flows from the spec itself.”

How It Works: Plugins, No Code

The system operates entirely through Claude Code Plugins. Advertisers describe their campaign in a chat interface; Claude interprets the intent, maps it to the correct API endpoints, and executes the calls securely. The plugin handles authentication, validation, and error recovery autonomously.

Since no custom plugin code was written, Spotify avoided months of development. The entire proof-of-concept came together in days, according to team sources. The tool also supports multi-turn conversations—advertisers can refine their campaign iteratively (“Target age 18-25 instead,” “Increase budget by 20%”) without restarting.

What This Means for Advertisers and Developers

For advertisers, the tool could democratize access to premium ad inventory. A small business owner could launch a Spotify campaign without hiring a developer or learning code. Campaign management becomes as easy as sending a message.

For developers, the approach signals a new paradigm: APIs that explain themselves to AI, enabling natural language interfaces without custom middleware. Spotify's success suggests that any API with a well-written OpenAPI spec could be turned into a conversational interface using similar AI plugins.

“This is the beginning of the ‘API-as-Conversation’ era,” said Chen. “We're already exploring ways to let users ask questions like ‘How did my last campaign perform?’ and get human-readable analysis.”

Industry analysts note that the technology could reduce time-to-campaign by up to 80% and lower the cost of ad tech integrations. However, concerns about AI reliability and API misuse remain.

Future Plans and Availability

Spotify plans to roll out the conversational interface to select beta partners in Q3 2025. The company also intends to open-source parts of the plugin methodology to help other platforms build similar tools.

The engineering team emphasizes that no proprietary data or code was exposed during development—the plugin operates within Spotify's secure environment. This approach could become a blueprint for other tech companies looking to simplify their APIs.

Expert Reaction

“Using AI to bridge the gap between complex APIs and end users is a natural evolution,” said Dr. Maria Lopez, an AI researcher at MIT. “Spotify's implementation is clever because it leverages existing specs without needing to train a custom model.”

Ad tech veteran Tom Williams commented: “This could be the ‘iPhone moment’ for programmatic advertising. If you can just say what you want and it happens, the whole industry changes.”

The project reinforces Spotify's commitment to innovation in advertising, following earlier moves into podcast ads and audio-first formats.

For more on Spotify's engineering work, see their blog post: original article.