Comment by mfrye0
I'm working on Plaid / Perplexity for business data.
The basic idea is that integrating business data into a B2B app or AI agent process is a pain. On one side there's web data providers (Clearbit, Apollo, ZoomInfo) then on the other, 150 year old legacy providers based on government data (D&B, Factset, Moody's, etc). You'd be surprised to learn how much manual work is still happening - teams of people just manually researching business entities all day.
At a high level, we're building out a series of composable deep research APIs. It's built on a business graph powered by integrations to global government registrars and a realtime web search index. Our government data index is 265M records so far.
We're still pretty early and working with enterprise design partners for finance and compliance use cases. Open to any thoughts or feedback.