Which episode did they talk about that?
A podcast is hundreds of hours of spoken audio that nobody can search. Taffy has already indexed 2,511 videos and their comments from 14 AI-builder channels, including Y Combinator, a16z and Greg Isenberg. Ask which episode a topic came up in and jump straight to the timestamp.
2 questions free with no account, then 5 a day signed in. Or read the research first.
Works on the shows in Taffy's AI-builder library.
The archive grows. It never gets more findable.
Three hundred episodes, four hours each. Everything worth knowing is in there somewhere and there is no way to search for it. Not for you and not for the listener who swears you covered this once.
"Which episode was that in?"
A guest or a study or a story you half-remember. You know you said it. You cannot find it. Neither can the listener asking in the comments, so the answer stays lost even though it exists.
The back catalog just sits there
Every great segment you ever recorded is a clip. It is also a newsletter and a shownotes link and a sequel. But repurposing means re-listening and re-listening to three hundred hours is not a plan.
The audience is telling you what to make next
Guest requests and topic requests. The questions that keep coming back. All of it sits in the comments across hundreds of episodes, where nobody reads it in one place.
Ask the shows Taffy already indexed.
Taffy has already indexed the episode transcripts and comments of the podcasts in its AI-builder library. You ask a question in plain language and get an answer with the episode and timestamp attached.
- Every episode's transcript, searched by meaning instead of exact words.
- The exact episode and timestamp for every answer, so you can jump straight there.
- Comments indexed alongside the transcripts.
- Covers the podcasts in the AI-builder library.
Five things a searchable archive unlocks
An indexed back catalog answers questions that hours of scrubbing never could.
Which episode mentioned it
Ask what the host said about a topic, a guest or a study across the whole show. Get back the episodes and the timestamps, ranked — not a list of titles to go read yourself.
Mine what you already recorded
Surface the strongest self-contained segments across every episode — the ones worth a short, a newsletter or a re-release. The archive stops being dead weight and starts being a content library.
"Which episode should I start with?"
The most common question a big show gets and the hardest to answer at scale. Point a listener's question at the index and hand back the exact episodes that cover what they asked about.
Research a show before you go on it
If the show is in the library, ask what the host already covered before you go on. Then bring a segment nobody else did.
What audiences keep requesting
Comments sit in the index next to every episode. Taffy's research issues read them for the guests and topics audiences keep asking for.
Search the words, not the title
Episode titles are marketing. What you need sits forty minutes into an episode with an unrelated title. Taffy searches the spoken content, so you find the segment even when the title gives no hint.
- "Where did the host explain their morning routine?"
- "Which guest talked about raising a first fund?"
- "Find every time they recommended a book."
Chat with the whole show
Ask questions across every episode of the podcasts in the library. Each answer cites the episode and the timestamp.
- "Summarize everything they've said about pricing."
- "Which guests talked about hiring a first engineer?"
- "Where did the host explain how they got their first users?"
Who searches a podcast this way
- Hosts with a deep back catalog who want to repurpose old segments and answer "which episode" without re-listening.
- Listeners and researchers tracking what a favourite show has said about a topic across years of episodes.
- Producers and editors pulling clips and shownotes from long episodes on a schedule.
- Guests and bookers researching a show's archive and audience before an appearance.
Podcast search FAQ
YouTube and podcast apps only search episode titles and descriptions, not what was said inside an episode. To find which episode a topic came up in, the transcripts have to be indexed. Taffy has indexed every transcript of the podcasts in its AI-builder library and searches them by meaning. It returns the episode and the exact timestamp, so you can jump straight there instead of scrubbing through hours of audio.
Yes, if the show is in Taffy's library. Taffy covers a curated set of AI-builder channels, including podcasts such as Lenny's Podcast and My First Million, and adds channels as the research needs them.
No. Taffy indexes YouTube channels, and only the ones in its library. A show that never posts to YouTube has nothing for Taffy to index. For shows in the library, it does not matter whether an episode is full video or a static audio waveform, because Taffy reads the transcript.
Taffy indexes comments alongside the transcripts, and its research issues use them to find the questions audiences keep asking. When you ask the library a question, the answer comes from the transcripts and cites the episode and timestamp.
Those search show and episode titles, and sometimes descriptions. None of them search the spoken content of an episode. Taffy searches the transcripts of the shows in its library and cites the episode and timestamp for every answer.
Search the archives Taffy already indexed.
2 questions free with no account, then 5 a day signed in. Go Pro for $9 a month when you want every archive and no cap.