People data for podcasts
What’s in it
- Where every credit came from. Each host or guest credit records how we know, such as the show notes or the feed’s own cast list.
- Accuracy we measure. Every week a second, stronger AI re-reads a sample of episodes and scores the credits, and we fix the rules when it finds a mistake.
- Updated every night from the latest episodes.
- One permanent id per person across shows, so “everyone this guest has been on” is one lookup. Linked to Wikidata where the person has an entry.
- So far, more than 130,000 people and 870,000 appearances across 18,000 shows, and growing.
It is people data, not a podcast directory. It links to each episode, but does not include show notes, artwork or audio, which belong to their publishers. People can ask us to correct or remove their page under our corrections and removal policy, and API customers will apply those removals too.
Example
The shape of one person’s record, with made-up values and two appearances. The final format may differ.
{
"id": "pi_000123",
"name": "Dana Rivera",
"wikidata": "Q00000000",
"appearances": [
{ "show": "Example Interview Show", "episode": "Dana Rivera on her new book",
"date": "2026-10-14", "role": "guest", "source": "show notes" },
{ "show": "Dana Rivera's Weekly", "episode": "Episode 212",
"date": "2026-10-20", "role": "host", "source": "show notes" }
]
}
Pricing
Not on sale yet. We plan a free option for personal and research use, paid plans for products, and a full-dataset licence for larger teams. People on the early-access list hear first and help us set the prices.