Utterances like these should Just Work:
  • “Get me the HR leadership at Apple”
  • “Give me CHRO-type people at Apple”
  • “CEO at StayVista”
You send one naturalLanguage utterance (role, optionally at company). Arxena resolves synonyms and seniority server-side, maps std function/root/grade to LinkedIn Sales Navigator or Apollo filters when using those sources, then returns people plus a small resolved object. If the utterance has no company and you omit companyName / companyId / website, the API returns Please provide company name as well. A company name alone is resolved through companies ES → autocomplete → Google SERP domain → LinkedIn company id before search. A domain alone uses ES/autocomplete first, then Google {domain} linkedin company if that is only a stem guess.

​
One call (index)

curl -X POST \
  -H 'Authorization: Bearer YOUR_API_KEY_TOKEN' \
  -H 'Content-Type: application/json' \
  https://app.arxena.com/people-api/people/search \
  -d '{
    "naturalLanguage": "CHRO at Apple",
    "dataSource": "index",
    "limit": 10
  }'

​
Sales Navigator pipeline (Unipile pool)

Classifies the title → maps to Sales Nav function IDs + seniority → searches via the shared Unipile Sales Navigator pool → post-filters by taxonomy. Require one of:
SourceBody fields
Shared Sales Nav pool"dataSource": "pool"
Harvest Sales Nav"dataSource": "harvest"
Explicit Unipile account"dataSource": "unipile", "accountId": "..."
curl -X POST \
  -H 'Authorization: Bearer YOUR_API_KEY_TOKEN' \
  -H 'Content-Type: application/json' \
  https://app.arxena.com/people-api/people/search \
  -d '{
    "naturalLanguage": "CEO at StayVista",
    "dataSource": "pool",
    "limit": 10
  }'
Response includes:
  • resolved — std function / root / grade from the title
  • query.appliedFilters — Sales Nav functionIds + seniorities (or Apollo department/seniority when dataSource: "apollo")
  • items — people after taxonomy post-filter

​
Apollo path

Same endpoint with "dataSource": "apollo". Taxonomy maps to person_department_or_subdepartments (including master_* roots) and person_seniorities.

​
What you see back

  • resolved — what the phrase became (function root × grade + confidence)
  • items — filtered people at that company
Playground: Search people.