Radar turns podcasts into searchable, AI-readable data
A startup called Particle has built a search engine that transcribes more than 130,000 podcasts and makes spoken audio usable by AI agents. Hedge funds are already paying for it.
Key points
- Particle, founded by former Twitter engineers, launched Radar, a podcast search engine, on Wednesday.
- Radar transcribes more than 130,000 podcasts, adding 20,000 new episodes to its index every day.
- Hedge funds are the highest-volume API customers, integrating Radar directly into their research tools.
- Personal plans start at $29 a month per seat; business plans cost $399 a month for 20 seats.
- Particle plans to expand Radar beyond podcasts to YouTube videos and news clips.
Most of the conversation on the internet isn't written down. It's spoken: on podcasts, in interviews, across hundreds of thousands of audio episodes published every week. That content has been largely invisible to AI, and Particle wants to change that.
Particle, an AI startup founded by former Twitter engineers, launched Radar on Wednesday. The service transcribes podcast audio, identifies the people, brands and topics mentioned, and makes all of it searchable.
What does Radar actually do?
Radar goes beyond transcription: it tags entities so you can search a name and surface every podcast moment it appears in. Users can set up alerts delivered by email, Slack or webhook whenever a specific person or company is mentioned, with a timestamped clip attached. The clips are self-contained, readable without sitting through a full episode.
Radar covers all 200 of Apple's top-ranked podcasts across 135 subject categories. By Particle's own count, that makes it the largest transcribed podcast service currently available, with 20,000 episodes added daily.
Who is paying for this?
Hedge funds are the biggest customers right now. "Hedge funds have been the highest-volume customers that are directly integrating with the API," Particle co-founder and CEO Sara Beykpour told TechCrunch. Financial analysts spend considerable effort tracking what executives and market commentators say publicly, and podcasts have become a primary venue for those conversations. Our earlier story on Situational Awareness LP showed how aggressively some funds are building AI into their research pipelines, which helps explain the appetite here.
"Agents are generally blind to audio; they can't see it unless something or someone has transcribed it," Beykpour said. AI search platform Exa is already listed as a partner. Journalists and data resellers are also among paying users.
| Plan | Price | What you get |
|---|---|---|
| Personal | $29/month per seat | Web search, alerts, clip extraction |
| Business | $399/month | 20 seats, full feature set |
| API | Custom pricing | Programmatic access for developers and businesses |
Particle describes the API, a technical connection that lets other software pull Radar's data automatically, as the core product. An MCP, a newer standard that makes it easier for AI agents to call up outside data sources, is also available.
What does this mean for ordinary listeners?
Nothing changes about how podcasts sound or where you find them. Radar works in the background against content that is already public.
For professionals tracking brands or public figures, the upside is concrete: fewer hours of audio to monitor, and targeted alerts when relevant moments occur. Worth watching, though, is the accuracy question. Our 21 August story found that several popular transcription systems reproduce benchmark errors even when the audio says something different; a hedge fund acting on a misheard quote could face real consequences.
Particle plans to extend Radar to YouTube videos and broadcast news clips, widening the pool of spoken content that AI tools can read.
Common questions
Do podcast creators need to opt in?
Radar transcribes publicly available podcast audio, in the same way a search engine indexes public websites. No opt-in from creators is required under current practice.
Is my private audio at risk?
Radar indexes public podcast feeds only. Private or paywalled episodes not in a public RSS feed, the standard distribution format podcasts use, are not included.
Can I use Radar without a technical background?
Yes. The web interface lets anyone search podcasts, set up alerts and read extracted clips without writing any code. Developer tools are aimed at businesses building their own software on top.
