1. Approved use
Named users may use Sentinel Graph only for the healthcare legal, investigative, compliance, SIU, or program-integrity purpose approved in writing. Use must be lawful, proportionate, professionally supervised, and subject to independent human verification.
2. Required verification
- confirm identity and current status at the cited publisher;
- distinguish exact-NPI matches from name-and-state candidates;
- consider ordinary explanations for shared infrastructure;
- treat scores as review ordering, never as a probability or conclusion; and
- document independent reasoning before relying on a lead.
3. Prohibited adverse-decision use
No credit, employment, housing, insurance, eligibility, licensing, contracting, enforcement, referral, prosecution, or other adverse-decision use is permitted. Output may not be the sole basis for any action, and users may not describe it as a finding or accusation of fraud or wrongdoing.
4. Prohibited access and extraction
- no credential sharing or unnamed-user access;
- no scraping, crawling, bulk-download, bulk-export, automated extraction, or rate-limit evasion;
- no resale, redistribution, publication, data brokerage, or model training;
- no deanonymization of masked samples or attempts to access disabled or administrative surfaces; and
- no unlawful, misleading, harassing, discriminatory, or retaliatory use.
5. Sensitive information and AI
Do not submit PHI, medical records, confidential client information, protected case material, non-public investigative information, Social Security numbers, credentials, or secrets. Do not prompt AI-assisted features to ignore safety rules, invent evidence, make an accusation, or advise an adverse action. Verify AI-assisted text against the structured graph and cited sources.
6. Enforcement
Sentinel Graph LLC may rate-limit, suspend, or revoke access when it reasonably believes this policy has been violated or a security risk exists. Report suspected misuse or account compromise to probert@sentinelgraph.com.