Data Sources & Methodology

Public-source intelligence, rigorously structured

SentinelGraph works exclusively with publicly available records. Our methodology is transparent, our findings are independently verifiable, and our process does not depend on proprietary or privileged data sources.

Our public-source approach

We use only records that are publicly available — meaning they can be accessed by any member of the public through government websites, public databases, FOIA responses, or other lawful channels. This approach has three important advantages:

Transparency

Every finding can be traced to a specific, identified public source. There are no black-box algorithms, no proprietary scoring models, and no undisclosed data inputs.

Verifiability

Because every source is public, every finding can be independently verified by the recipient. This is essential for legal, regulatory, and compliance use cases where findings must be defensible.

Independence

We do not rely on data from the entities or individuals being analyzed. Our findings are based entirely on external, independently obtained records.

Why public-source methodology matters

Working exclusively with publicly available records is not a limitation — it is a deliberate design choice that strengthens the value of SentinelGraph findings in professional settings.

Transparency

Every source is identified and documented. There are no hidden inputs, proprietary scores, or undisclosed algorithms. Recipients can evaluate exactly how each finding was derived.

Portability

Because findings are built from public records, they can be shared across organizations, included in legal filings, and referenced in regulatory communications without sourcing restrictions.

Professional review settings

Attorneys, investigators, and compliance officers can independently verify each cited record. This makes SentinelGraph outputs suitable for use in legal briefings, compliance audits, and investigative referrals.

Record categories

SentinelGraph ingests and structures records from the following public-source categories:

Core Federal Sources

  • OIG-LEIE — List of Excluded Individuals/Entities (82,000+ records)
  • SAM.gov — System for Award Management exclusions
  • PECOS — Provider Enrollment, Chain, and Ownership System (2.5M+ enrollments)
  • CMS Ordering & Referring Registry (2M+ providers)
  • NPPES — National Plan and Provider Enumeration System

Facility & Enrichment Data

  • CMS Hospital Compare — general hospital information
  • CMS Nursing Home Compare — facility-level provider data
  • CMS Home Health Compare — home health agency data
  • CMS Dialysis Facility Compare — dialysis facility data
  • HRSA Health Center Sites — federally qualified health centers
  • CMS Hospital Ownership — beneficial ownership records (147,000+)
  • CMS Nursing Home Health Citations — enforcement events (418,000+)
  • CMS Physician/Clinician National File (1.4M+ providers)

Enforcement & Intelligence

  • CourtListener — Federal court records cross-referenced daily (7,000+ cases)
  • GDELT News Intelligence — global news coverage matched to leads (6,000+ articles)
  • HHS-OIG Enforcement Actions — settlements, indictments, and convictions
  • OpenSanctions — aggregated state Medicaid exclusion data

State & Corporate Records

  • State Medicaid exclusion lists (California, Texas, and expanding)
  • Colorado Corporate Registry (3M+ business entities)
  • Florida SunBiz Corporate Registry
  • Secretary of State filings across multiple jurisdictions
  • Corporate officer, director, and registered agent records

How findings are handled

Not every pattern detected by SentinelGraph becomes a finding delivered to a client. Our process includes multiple quality gates:

1

Automated detection

Entity resolution and pattern-matching algorithms identify potential control indicators across the data graph.

2

Confidence assessment

Each potential indicator is scored for source reliability, cross-source corroboration, temporal consistency, and entity resolution confidence.

3

Human review

A human analyst reviews every finding before it reaches a client. The analyst assesses context, confirms evidence chain integrity, and assigns a final confidence rating.

4

Structured delivery

Findings that pass human review are delivered as structured intelligence — with evidence citations, confidence ratings, and clear statements about what was observed, what it may indicate, and what remains unknown.

Evidence traceability

Every SentinelGraph finding includes a complete evidence chain. For each observation, we document:

The specific public record(s) supporting the observation

The source authority (e.g., state agency, federal database)

The date the record was obtained or last verified

The confidence level of the entity resolution (identity matching)

Any corroborating records from independent sources

The analyst's assessment of what the evidence supports and what it does not

Important limitations

SentinelGraph works with publicly available data. This means there are things we cannot see:

  • Private agreements, informal arrangements, and verbal understandings are not reflected in public records.
  • Corporate structures designed to obscure control may not be fully penetrable through public filings alone.
  • Records may be incomplete, outdated, or inconsistent across jurisdictions.
  • Entity resolution across data sources is probabilistic — name and address matching can produce both false positives and false negatives.

These limitations are inherent in any public-source intelligence methodology. We address them through cross-source corroboration, confidence rating, and human review — but we do not claim to see everything. Our findings are starting points for qualified professional review, not endpoints.

There are things public records cannot show on their own, including private agreements, unrecorded management relationships, or claims facts not reflected in public datasets. SentinelGraph outputs should therefore be treated as structured starting points for qualified review, not final endpoints.

Request a confidential briefing

If you work in healthcare compliance, fraud investigation, or legal oversight — and you want to understand how exclusion-linked control risk may affect entities in your portfolio — SentinelGraph may be able to help.

Request a Briefing

Important disclaimer: SentinelGraph provides structured intelligence based on publicly available data. It does not provide legal advice, conduct investigations, or make determinations of fraud, liability, or regulatory violation. All findings are preliminary, evidence-traceable observations intended to support — not replace — qualified legal, investigative, or compliance review. Use of this service does not create an attorney-client relationship or any professional engagement unless separately agreed in writing.