Radar makes podcasts searchable for AI agents
Particle has launched Radar, a platform that transcribes and indexes more than 130,000 podcasts. Its goal is to make audio conversations searchable and available to AI agents through an API and MCP.
Summary
Particle has launched Radar, a podcast intelligence platform that turns audio conversations into searchable material. According to the company, the service transcribes and analyses more than 130,000 podcasts, making it possible to find quotes, topics and specific references without listening to entire episodes.
In practice
Radar adds speaker labels and metadata about the people, companies, brands, products and topics discussed. It can also send email, Slack or webhook alerts when selected terms, guests or subjects appear in a podcast. Its API and MCP integration are aimed at AI agents and businesses that want to query this data programmatically.
Context
Most search systems for AI agents work primarily with text available on the web. Particle is trying to build a comparable layer for audio through transcription, indexing and timestamped clip extraction. The company says it adds around 20,000 episodes to its index each day and plans to expand beyond podcasts to formats such as YouTube videos and news clips.
Why it matters
- It could reduce the time needed to locate information in long-form conversations.
- It creates a new data source for brand monitoring, research and market analysis.
- Its usefulness will depend on transcription quality, catalogue coverage and accurate interpretation of context.