Find expert commentary

Goal: hear from people close to the subject — operators running the infrastructure, researchers covering the science — rather than generalist commentary.

Filter Search by what the speaker does professionally:

1
{
2
  "query": "AI inference infrastructure bottlenecks",
3
  "speaker_roles": ["operator", "researcher"],
4
  "include_quotes": true
5
}

With Ask, name the audience in the question and the agent applies the filter itself:

1
{"query": "What are operators saying about where the economics accrue in autonomous driving?"}

Look for: the speaker’s role and title (speaker_role, speaker_title) — an operator describing their own cost structure is a different kind of evidence than an investor’s read of it. Weigh first-hand experience, and check whether practitioners agree with each other before treating one voice as the field’s view.

A pattern across all recipes: use Ask to open an investigation, then Search to pull the specific evidence sets you want to work with. The insight objects that come back are the same either way — see The Data Model.