12 guests, 1,000+ sources, 6 hours
Every interview a guest has ever given, every gap nobody's explored — before the host walks on set.
I researched 12 guests in 6 hours for a podcast shoot — every interview they've ever given, every gap nobody's explored, questions ranked by clip potential.
The content was better than episodes we'd spent a week prepping.
The problem
You can't ask a question nobody's asked if you don't know what everyone's already asked.
Most podcast producers don't. They Google the guest, skim a Wikipedia page, maybe watch one previous interview if they're feeling diligent. Then they walk in and ask "tell me your story" — the same question 50 other hosts asked. The guest clicks into autopilot, gives the rehearsed version, and everyone pretends this was a conversation.
A guest who's done 40 interviews has answered hundreds of questions publicly. Somewhere in there are gaps — things they've never been asked, recent work that hasn't been covered. But finding those gaps means consuming all 40 interviews. No producer has time for that.
What the system does
For every guest on The Masoom Minawala Show, it consumes 100s of sources — not just their interviews, but their books, academic work, social media, news coverage, even the people around them. Family, co-founders, collaborators. It runs in spirals: every source it reads surfaces new leads, which surface new sources, until there's nothing left to find.
The output is an 80-page brief. What they've said everywhere, what they've never been asked, where they've contradicted themselves across interviews, which stories have emotional weight, which angles map to our audience. Every claim fact-checked before Masoom walks on set.
12 guests. 1,000+ sources. 6 hours. A team of researchers couldn't match the depth in a month.
Most producers think they need better questions. They need better information. The questions write themselves after that.