Particle, the AI newsreader startup created by former Twitter engineers, is expanding into podcast intelligence with a new product called Radar. The platform transcribes audio, interprets context, and turns spoken conversations into searchable data for people and AI systems alike.
Radar is designed to do more than convert speech to text. It identifies key quotes, highlights important moments, and adds rich metadata such as speakers, entities, topics, and timestamps. The system now indexes more than 130,000 podcasts, including the Apple Top 200 across 135 categories, and adds around 20,000 episodes each day.
According to Particle co-founder and CEO Sara Beykpour, the strongest demand has come from hedge funds, AI search platforms, and data resellers. Radar also supports API access, making it easier for AI agents and businesses to integrate podcast intelligence directly into their workflows.
Users can set alerts for specific names, brands, or themes, delivered through email, Slack, or webhook. Radar can also surface self-contained clips with timestamps, track ad mentions, and provide tools for ranking, audience estimation, and sponsorship analysis.
Particle says the long-term goal is to extend this audio intelligence layer beyond podcasts to formats such as YouTube videos and news clips. The shift reflects a broader move toward making spoken media as searchable and usable as text, opening new possibilities for discovery, analysis, and automation in the digital future.
