match_person

Resolve people via OnFire’s Matchbox2 engine. Vector search and AI matching handle “Lana P. at Snowflake” or just an email alias, and return the verified LinkedIn profile, job title, and current company.

When to use it

• You have a person but no LinkedIn URL and want to score them with ai_prospecting.
• A list of CRM contacts needs LinkedIn URLs before contact_data_enrichment.
• You want to verify the person still works at the company.

Skip this if you already have a clean person LinkedIn URL and the user isn’t asking you to verify identity.

Inputs

Parameter	Type	Required	Description
entities	object[]	Yes	Up to 100 person descriptors.

Each entity can include any combination of:

• person_full_name — e.g. "John Smith".
• person_email — e.g. "john@acme.com".
• company_name — e.g. "Acme Inc".
• job_title — improves match accuracy.
• company_linkedin_url — improves match accuracy.
• company_website — improves match accuracy.

Minimum: at least one of person_full_name or person_email. If using name without email, company_name is required.

Returns

JSON with total_count, matched_count, and a results array in the same order as the inputs. Each matched record carries:

• matched (boolean)
• linkedin_url, full_name, job_title
• company_name, company_linkedin_url, company_website
• match_score, match_type

Example

{
  "entities": [
    { "person_full_name": "John Smith", "company_name": "Acme Inc" },
    { "person_full_name": "Jane Doe", "person_email": "jane@techcorp.com", "company_name": "TechCorp" },
    { "person_email": "ceo@startup.io" }
  ]
}

Notes

• Batch up to 100 entities per call.
• A successful match usually returns the company LinkedIn URL too — reuse it for downstream ai_prospecting(action="get_prospect") calls.