Particle Launches Radar to Unlock Podcast Data for AI Agents

AI startup Particle has launched Radar, a podcast search engine and intelligence platform designed to make audio content accessible to AI agents. The move marks a strategic pivot for the company, founded by former Twitter engineers, as it shifts focus from its original AI news-reader application toward audio-intelligence infrastructure.

Radar currently transcribes and analyzes more than 130,000 podcasts, a catalog that includes every podcast in the Apple Top 200 across 135 different verticals. To maintain the freshness of its index, the platform adds approximately 20,000 new episodes daily. For businesses and developers, Particle provides programmatic access through an API and the Model Context Protocol (MCP). Pricing for the platform begins at $29 per month per seat, while a business plan for 20 seats is available for $399 per month; API pricing is handled on a custom basis. The company has also partnered with AI search provider Exa.

The Pivot to Audio Intelligence

Particle's transition was driven by internal discovery. While operating its AI-powered news reader, the team found that the internal tool they had built to source podcast clips was significantly more valuable than their consumer-facing product. This realization led the company to pivot toward a dedicated API.

This shift addresses a critical blind spot in the current AI ecosystem. Most AI agents rely on crawling text-based web data, leaving them unable to process spoken content. As Sara Beykpour, Particle co-founder and CEO, noted, "Agents are generally blind to audio; they can’t see it unless something or someone has transcribed it."

Unlocking 'Dark' Data

By converting massive volumes of spoken audio into structured, searchable data—complete with entity recognition and speaker labels—Radar transforms a previously "dark" data source into actionable intelligence. This capability is particularly valuable for financial analysts and AI agents who require real-time monitoring of corporate mentions, advertising trends, and political bias across the podcasting medium at scale.

According to TechCrunch, hedge funds have already emerged as the highest-volume customers directly integrating with the Radar API, highlighting the immediate demand for structured audio data in high-stakes financial environments.

The Path Forward

As AI agents become more integrated into professional workflows, the ability to ingest non-textual data will be a key differentiator. The industry will now be watching to see if other audio formats are integrated into Radar's index and how the adoption of the Model Context Protocol accelerates the integration of podcast intelligence into broader AI agent frameworks.