For AI agents: llms.txt.

Sources & confidence

Match score

The match score estimates how well a candidate fits the query. It can use compatible names, handles, locations, employers, profile cross-links, photos, and verified account signals.

A high score means the evidence fits well; it does not prove identity. A low score can still be the right person when the public footprint is sparse.

Strong

Multiple independent signals agree and no important contradiction is visible.

Possible

Some useful signals agree, but the candidate still needs manual checking.

Weak

The result is mainly a name or handle resemblance with little corroboration.

Conflicts and estimates

Some facts are estimates, especially visual analysis, inferred locations, company financials, or profile relationships. DeepSearch labels confidence where it can, but the safest interpretation is to separate direct facts from model-assisted conclusions.

Always verify consequential facts
Do not rely on a match score, age estimate, relationship label, location inference, or company estimate as the sole basis for a legal, financial, employment, housing, or safety decision.
Skip to content