Particle, an AI newsreader startup founded by former Twitter engineers, announced Radar, a podcast search engine that transcribes audio and indexes spoken content to make it searchable. The service now indexes more than 130,000 podcasts and adds about 20,000 episodes daily, including speaker labels and metadata that identify people, companies, brands, products and topics. Radar can track mentions of these entities and send alerts via email, Slack or webhook, with filters to specify guests or subjects. Users can retrieve self‑contained clips with timestamps for listening or reading. The company offers the product through a web interface and an API with a meeting protocol (MCP) for programmatic access. Pricing is $29 per month per seat, with a $399 plan for businesses that includes 20 seats; API pricing is custom. Radar’s partners include the AI search provider Exa. The company said its vision is to provide audio intelligence through an API, addressing a gap where most AI agents lack the ability to understand spoken media. Future plans include expanding the service to support other audio formats such as YouTube videos and news clips.