Capability 01 · Know the account
HCP Intelligence
Turn a named HCP or HCO research request into a structured public-source profile that an account team can inspect—not a score that hides its evidence.
The problem¶
Account preparation often spreads identity clues, affiliations, publications, trials, grants, and web findings across disconnected searches. A namesake or stale affiliation can then travel into a briefing without its uncertainty.
Objective¶
Create one reviewable HCP or HCO profile in which factual claims retain source URLs, access dates, confidence, and disambiguation notes. The objective is better-prepared human research and account discussion—not automated targeting.
How it helps¶
- Searches supported public sources including PubMed, ClinicalTrials.gov, ORCID, NIH RePORTER, and configured web providers.
- Keeps account identity, evidence-backed claims, source coverage, confidence, and profile completeness together.
- Supports CSV batch work with an exact dry-run scope gate and reviewable JSON, CSV, and manifest artifacts.
How the plugin works¶
Account identity and scope
A governed account ID or CSV with name, country, account type, specialty, and institution clues.
Account and public-source research
list_accountsget_accountupdate_accountsearch_publicationssearch_guidelinessearch_clinical_trialssearch_grantssearch_orcidsearch_congressessearch_hcp_websearch_hco_web
Cited profile and batch record
Evidence-backed profile JSON; batch work also writes a summary CSV and manifest with completion status.
A three-step workflow¶
Resolve the subject
Supply a governed ID, name, country, account type, and useful affiliation or specialty clues.
Gather public evidence
Search the relevant sources and preserve URLs, dates, provider coverage, and namesake evidence.
Inspect the profile
Check identity, confidence, missing coverage, recency, and every partial or failed record before use.
Sample input¶
The downloadable CSV uses the pinned batch header: id,name,specialty,country,account_type,institution. Its two rows
are explicitly fictional and show one HCP and one HCO without exposing a real person or customer.
Interpreted sample output¶
The representative JSON excerpt gives the fictional title medium confidence because only one simulated
institutional source supports it. Its profile_completeness is 0.42 and the review flags call out absent
corroboration. That is a reason to continue verification, not a measure of HCP value or commercial priority.
Download representative output JSON
Business value indicators¶
| Indicator | What a business reviewer learns |
|---|---|
| Identity exceptions | Where namesake or affiliation conflicts still need resolution |
| Claim source coverage | Which material profile statements have linked evidence |
| Source status and recency | Which providers succeeded and when evidence was accessed |
| Partial or failed account count | Where a batch is incomplete and must not be treated as finished |
| Profile completeness | A review aid for missing fields—not a quality or influence score |
Boundaries and human review¶
Public-source results may be incomplete, stale, or refer to a namesake. Confirm identity, source authority, data rights, privacy, consent, retention, and operational relevance. If optional synthesis is used, confirm the selected provider is approved for the submitted account fields and evidence. This public-beta capability does not determine whether an HCP should be targeted; qualified people remain accountable for that decision.
Get started¶
Inspect the small fictional files first, then use the canonical pinned guide for current inputs, provider gates, and batch completion rules.