AI Chats on Podcast Episodes
Chat with an AI agent about any podcast episode you've listened to. Ask questions, get quotes, explore tangents.
Who this is for: Podcast listeners, researchers, students, anyone who wants to go deeper on an episode
How it works with feed.works
Once a podcast episode is transcribed, the transcript becomes context that an AI agent can read, reference, and reason about. Connect your feed.works library to an MCP endpoint and any compatible agent — Claude, ChatGPT, a custom assistant — can answer questions about what was said, who said it, and what it means. OpenClaw-style agents can now chat with you about your podcasts the way a friend who also listened would — except this friend has perfect recall and never gets the quote wrong.
Feeds you'd subscribe to
- Hard Fork (NYT)
- Acquired
- Lex Fridman Podcast
- The All-In Podcast
- Any podcast you already subscribe to
Workers you'd use
How you'd use all this data
- Transcribe worker processes every episode automatically — the transcript is the foundation
- Show Notes worker extracts chapters, key quotes, and named entities for richer context
- Connect your library via MCP endpoint so your AI agent can access the full transcript
- Ask natural-language questions: 'What did the guest say about regulation?' or 'Summarize the debate about open source'
- Agent cites exact quotes with timestamps — click to jump to that moment in the audio
- Cross-episode questions work too: 'What has Chamath said about interest rates across the last 5 episodes?'
- Entity extraction gives the agent structured knowledge about people, companies, and topics mentioned
Ready to try AI Chats on Podcast Episodes?
Start with the free plan. Subscribe to feeds, assign workers, and see what your data looks like enriched.