Radar Turns Podcasts Into Searchable Data for AI Agents

Particle, the AI newsreader startup founded by former Twitter engineers, is expanding beyond news summaries with a new platform designed to make podcast conversations easier for people and artificial intelligence systems to search.

The company has launched Radar, a podcast search engine that transcribes audio, identifies important information and extracts relevant quotes and highlights from spoken conversations. The platform is designed to give businesses and AI agents access to information that has traditionally remained difficult to analyse because it exists only in audio form.

Particle CEO and co-founder Sara Beykpour said hedge funds have become among the highest-volume users integrating Radar via its API. Journalists, researchers, AI search platforms and data providers are also among the groups interested in accessing podcast intelligence.

Making Audio Accessible to AI Systems

The idea behind Radar came from a feature inside Particle’s news application that used its technology to discover useful podcast clips and display them alongside related news stories. The company later decided to expand that capability into a standalone platform focused on audio intelligence.

Beykpour said many AI systems and services are designed to search web-based text but cannot directly access information contained in audio unless it has already been transcribed. Radar is designed to provide that missing layer by making spoken content available through searchable data.

The platform currently transcribes more than 130,000 podcasts, including Apple’s top 200 podcasts across 135 categories. Particle said Radar adds about 20,000 new episodes to its index every day.

Podcast Tracking, Alerts and AI Access

Radar’s transcripts include speaker identification and metadata that allows the system to understand references to people, companies, brands, products and topics discussed during episodes.

Users can track specific mentions across podcasts and receive alerts when selected topics or entities appear. Alerts can be delivered via email, Slack, or webhook integrations, with filters that allow users to narrow results by guests, subjects, or podcast rankings.

The platform can also identify podcast clips with timestamps, allowing users to listen to specific moments while reviewing the related transcript. In addition, Radar tracks topics, advertisements, ratings, reviews and other episode information.

Particle has also built a podcast advertising search tool that can identify episodes where companies appear as sponsors and track those mentions over time. Additional features include tools for political bias analysis, rankings data, audience estimates, sponsorship information, and brand suitability.

While Radar is available through a web interface, the company says its API and MCP integration are the core products, allowing AI agents and businesses to access podcast intelligence programmatically. Radar also works with partners including Exa, an AI search API provider.

Radar is priced at $29 per month per seat, while businesses can purchase a $399 monthly plan that includes 20 seats. API pricing is customised based on usage requirements.

Particle plans to expand Radar beyond podcasts in the future, with support for additional audio sources such as YouTube videos and news clips.