Podcast and newsletter transcripts for your agents

We transcribe the podcasts and read the newsletters, and make them searchable. Set up alerts, or hand it all to your AI agent over MCP or API.

14 day free trial. Cancel anytime.

Brand and consumer intelligence

Brand health, product and trend signals, and who is talking about a brand before search and social pick it up.

“Chipotle, the interesting thing, I've stopped going there because I think quality is down. And I know I'm not alone there.”
Lou Whiteman, Motley Fool Hidden Gems Investing, 18 September 2026

Reputation and narrative intelligence

Where a story started, which named voices carried it, and how the framing changed along the way.

“The trial is expected to run about three months, but the reputational battle will be longer and harder to salvage.”
Ivy Yang, Calling the Shots, 22 September 2026

The data

Every newsletter issue and podcast episode is one record, with the same fields whichever it is. We enrich each one with the people's contact details, who the sponsors are, and what topics it covers. Archives go back as far as each publisher's feed does. Benedict Evans goes back to 2013, Acquired to 2015.

A sample record

One Dwarkesh Podcast episode, from three API calls: the episode, the first turns of its transcript, and the show. Long text is cut short here.

Speakers have their real names, and ad reads are labelled apart from the conversation. Those are the two things your pipeline would otherwise have to build.

Freshness

New issues and episodes are collected continuously as they're published, so the data stays fresh rather than arriving in a daily or weekly batch. Podcast transcripts follow once the episode has been transcribed.

{
  "post": {
    "slug": "z2cm-e6rd-9w8e",
    "title": "Noam Brown – Agent swarms, alignment, & recursive self-improvement",
    "link": "https://www.dwarkesh.com/p/noam-brown",
    "published_at": "2026-09-17T15:38:13Z",
    "source": {
      "slug": "jnvz-3qo7-i59c",
      "name": "Dwarkesh Podcast"
    },
    "transcript_status": "fetched",
    "people": [
      {
        "slug": "ofgm-9vx0-95yu",
        "name": "Dwarkesh Patel",
        "role": "host"
      },
      {
        "slug": "pefg-49b1-x7v2",
        "name": "Noam Brown",
        "role": "guest"
      }
    ],
    "topics": [
      {
        "slug": "drn6-87w4-0qv3",
        "name": "recursive self-improvement"
      },
      {
        "slug": "fxyi-6kds-kw9m",
        "name": "ai alignment"
      }
    ],
    "sponsors": [
      {
        "slug": "6cen-imru-gxei",
        "name": "Jane Street"
      },
      {
        "slug": "a3ai-psrg-dwzf",
        "name": "Antithesis"
      }
    ]
  },
  "transcript": {
    "total": 82,
    "has_speaker_diarization": true,
    "items": [
      {
        "speaker": "Dwarkesh Patel",        "kind": "speaker",
        "text": "Today, I'm chatting with Noam Brown, who is a researcher at OpenAI. He was one of the foundational contributors to what became O1 and the reasoning models, and now he's working on …"
      },
      {
        "speaker": "Noam Brown",        "kind": "speaker",
        "text": "So the way I think about it, when you plot the performance of these reasoning models with test time compute on the X axis and performance on basically any reasoning benchmark on the Y axis, …"
      },
      {
        "speaker": "Dwarkesh Patel",        "kind": "ad",        "text": "Suppose you need to do a major back-end refactor. Getting assurance that you didn't introduce any new bugs could take weeks of writing an extensive battery of tests, potentially more time …"
      }
    ]
  },
  "source": {
    "slug": "jnvz-3qo7-i59c",
    "name": "Dwarkesh Podcast",
    "type": "podcast",
    "authority_score": 70,
    "liveness": "active",
    "first_post_published_at": "2020-05-22T18:31:47Z",
    "posts_count": 182
  }
}

Why podcasts and newsletters

Search results and social feeds are filling up with generated content. People are moving their real opinions to spaces they chose: a newsletter they subscribed to, a podcast hosted by someone they trust.

58%

of Americans 12+ listen to a podcast every month

Edison Research, Infinite Dial 2026

23x

more trustworthy than social media, according to podcast listeners

The Podcast Study 2026

52%

named a podcast host as the influencer whose opinion matters most, against 24% for social media influencers

The Podcast Study 2026

That's where the early signals and real opinions show up now, and it's content a normal search index doesn't reach.

Delivery

API
Query by source, person, topic or sponsor.
MCP
Plug the corpus straight into an agent.
User interface
Search and browse the corpus directly, no integration needed.

License

  • Build on it and sell what you build.
  • Derived data stays yours when the contract ends.
  • Creator opt-outs are honored and takedowns are passed through to you.

Look at the data with us

On the call we look at the data together, for the companies and topics your customers care about.

Book a meeting