Particle, an AI newsreader startup founded by former Twitter engineers, is shifting toward podcast intelligence.
The company has launched Radar, a search engine for spoken conversations inside podcasts.
Radar does more than transcribe audio. It analyzes conversations and identifies important quotes, highlights, people, companies, brands, products, and topics.
Hedge Funds Are Using Radar
Particle co-founder and CEO Sara Beykpour said hedge funds now generate the most Radar API traffic.
Journalists and researchers can also use the service. Other customers include AI search platforms and data resellers.
AI search API provider Exa also works with Radar as a partner.
Particle originally developed the technology for its newsreader app. The app used the API to find useful podcast clips and display them with related news stories.
The company later saw an opportunity to turn the technology into a standalone product.
As AI agents gained popularity, Particle shifted its focus toward an API for audio intelligence.
Beykpour said most AI agents still search text across the web. However, these systems often miss audio content unless someone transcribes it first.
Radar Has Indexed More Than 130,000 Podcasts
Radar currently transcribes more than 130,000 podcasts.
Particle claims that this makes Radar the largest transcribed podcast service available.
The database includes all Apple Top 200 podcasts across 135 categories. Radar also adds around 20,000 new episodes every day.
Each transcript includes speaker labels and detailed metadata.
Radar can also identify the people, companies, brands, products, and topics mentioned in each episode.
Users can track these entities and receive alerts when they appear.
The service can send instant notifications or daily and weekly summaries. Users can receive alerts through email, Slack, or webhooks.
Filters provide more control over searches. For example, users can track a specific guest discussing a particular topic.
They can also limit searches to top-ranked podcasts.
Search Quotes, Clips, and Podcast Ads
Radar can find relevant clips and provide exact timestamps.
Users can then listen to the original audio or read the corresponding transcript.
Particle also selects notable clips for people who do not want to listen to an entire episode.
The platform tracks topics, mentions, listener ratings, reviews, advertisements, and other podcast data.
Radar also offers a dedicated podcast advertising search engine.
The tool can find episodes where a company placed an advertisement. It can then track those appearances over time.
Particle sees several other potential uses for the technology.
These include political bias analysis, podcast chart rankings, audience estimates, sponsorship research, and brand suitability tools.
Radar Offers API and MCP Access
Users can access Radar through a web interface. However, Particle sees its API and MCP as the main products.
These tools allow businesses and AI agents to access podcast intelligence programmatically.
Radar costs $29 per month for each seat.
The business plan costs $399 per month for 20 seats. Particle customizes API pricing based on each customer’s requirements.
Particle Plans Wider Audio Expansion
Particle plans to expand Radar beyond podcasts.
The company could eventually cover other audio and video sources. YouTube videos and news clips are among the potential additions.
The expansion could give AI agents broader access to information that exists outside traditional web text.
For now, Particle is focusing on making podcast conversations easier to search, track, and analyze.
