Quinta-feira, 27 de agosto de 2026

Aube.

As notícias do progresso
ImplementadoFonte única

Radar indexes 130,000 podcasts for AI agents

Idiomas deste artigo
Original · ENFR

Texto original em inglês. 2 idiomas disponíveis, o seu acrescenta-se com um clique.

A podcast episode can now be treated less like a long recording and more like a searchable database. Particle, the AI newsreader startup founded by former Twitter engineers, has launched Radar, a service that indexes more than 130,000 podcasts and makes their spoken conversations available to people and AI agents.

The service transcribes audio, labels speakers and identifies the entities being discussed — people, companies, brands, products and topics. It adds 20,000 episodes daily, covering all Apple Top 200 podcasts across its 135 verticals, according to Particle. Users can search for quotes and pull out self-contained clips with timestamps, so they can read a passage and listen to the original moment without going through an entire episode.

Radar also watches what changes over time. It can track mentions of a person, company or topic and send an alert when they appear, either immediately or in a daily or weekly digest. Filters can narrow those alerts to particular guests, subjects or top podcasts. A separate advertising search can find episodes where a company appears and follow that pattern over time. Particle also offers tools for political-bias analysis, chart rankings, audience-size estimates, sponsorship data and brand suitability.

The practical shift is aimed at software that can search websites but cannot directly interpret audio. Radar's API and MCP connection let AI agents and other businesses query the podcast index programmatically; Exa, a search API provider for AI agents, is among its partners. Sara Beykpour, Particle's co-founder and CEO, told TechCrunch that hedge funds are the highest-volume customers directly integrating the API, while journalists, researchers, AI search platforms and data resellers are other potential users.

So what changes in practice? A researcher can locate a relevant passage instead of scanning hours of audio, and a monitoring system can flag a new mention as it happens or collect it for later review. Radar costs $29 a month per seat, with a $399-per-month business plan covering 20 seats; API customers receive custom pricing. Particle plans to expand beyond podcasts to YouTube videos and news clips, but no independent assessment of Radar's accuracy has been published in the material available here.

130,000 podcastsPodcasts transcribed and indexed by Radar

Fontes — ler os originais(hora de Paris)

TechCrunchEN
0000

Para ler a seguir

Comentários

A carregar a conversa…

Inicie sessão para escrever um comentário. Iniciar sessão