Particle, an AI newsreader startup founded by a former Twitter engineer, is shifting its focus to a potentially more lucrative idea: indexing conversations buried in podcasts and making them discoverable. The company on Wednesday introduced Radar, a podcast search engine that not only transcribes podcast audio, but also understands its meaning and extracts important quotes and highlights.
Sara Beykpour, co-founder and CEO of Particle, explains that the solution has business potential, having already received interest from hedge funds looking for data that agents can’t see.
“Hedge funds are our highest volume customers that are directly integrating with our API,” Beykpour told TechCrunch. Journalists and researchers can also use the tool, but other high-paying customers include AI search platforms and data resellers. (For example, Exa, a search API provider for AI agents, is one of Radar’s partners.)
The idea itself came from one of the most popular features of the Particle news reading app. The app used an API to retrieve interesting podcast clips and include them in the app’s feed along with related news articles.
The team at Particle recognized the value of the product, but also realized that it was somewhat captive to a news reader. As the movement around AI agents began to gain momentum, the company decided to pivot and focus on building APIs for its podcast intelligence product.
“Our vision is really to bring all the new media intelligence and all the audio intelligence into that API. One of the reasons this is an interesting space is that most API agents and services are crawling the web and focusing on text. We’re bringing audio to that layer,” Beykpour said. “Agents are usually blind to audio; they can’t see it unless something or someone transcribes it.”
Using Radar, the company has transcribed over 130,000 podcasts, making it the largest podcast service in existence. It includes all Apple Top 200 podcasts across 135 subject areas and adds 20,000 episodes to Radar’s index every day.
Podcast transcriptions include speaker labels and rich metadata because Radar understands the entities being discussed (people, companies, brands, products, topics).
You can also track mentions of these entities throughout your podcast and send alerts when mentions occur or as a daily or weekly digest.
Alerts can be delivered via email, Slack, or webhooks, and can be customized using filters. These allow users to, for example, configure Radar to only send alerts when certain guests appear and discuss certain topics. You can also narrow your search in other ways, such as limiting it to only the top podcasts.
Radar can extract relevant, self-contained clips with timestamps, so users can listen to and read comments.
“We pre-select clips that are noteworthy, so if you can’t listen to the entire podcast and don’t want to read the synopsis, this is the best way to understand what’s going on in that podcast,” Beykpour said.
Radar can also track topics mentioned within the podcast, who or what was mentioned and when, listener ratings and reviews, episode ads, and more. There’s also a dedicated podcast advertising search engine that lets you search for all episodes in which a particular company advertises and track its trends over time. This feature has further monetization potential, as well as other tools that provide political bias analysis, chart ranking data, audience estimation, sponsorship data, and brand suitability.
All of this is available through Radar’s web interface, but the real product is the API and MCP that make this same intelligence programmatically available to AI agents and other companies.
Radar is priced at $29 per seat per month, with a 20-seat plan for $399 per month for businesses. API users can set custom prices based on their needs.
In the future, Radar plans to expand its service beyond podcasts and support other forms of audio, such as YouTube videos and news clips.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
