Skip to main content
Exa is a search engine built for AI research. Rippit uses it in two places: as the default provider for web search, and as one of three providers for people search. For how research tables work overall, see How research tables work.

Setup

None. Exa is built into Rippit, so you don’t need an Exa account or API key. Its usage is included in the cost Rippit reports for each run.

Web search with Exa

Exa is the default web search provider. It searches by meaning rather than exact keywords, so a brief like “companies announcing layoffs in their support team” finds relevant pages even when they don’t use those words. If you need a cheaper option or want to compare coverage, ask in the table’s chat to switch providers. Rippit also offers Parallel (similar quality, page excerpts and dates), Ceramic (much cheaper keyword search, snippets only, no publish dates), and DeepSeek (cheaper, with page content and dates). Changing the provider marks older samples as out of date.

People search with Exa

Exa searches structured professional profiles that match a natural-language description of who you want: role, company, location, and so on. Put those criteria in the brief, and Rippit turns them into searches and checks each profile against the brief before keeping it. The people search provider is fixed for a table after its first live run. Samples can compare providers before then. To change it later, create a new table.

Good fits

  • Tracking news, launches, or reviews about the companies in one of your tables.
  • Collecting long-form pages you want to classify or summarize with an AI table.
  • Building a list of people in a role or industry when you don’t need contact details yet.