Validation Case Study

How SentinelGraph surfaced a network later charged in a $50M+ federal healthcare fraud takedown

On April 2, 2026, the U.S. Department of Justice announced a Southern California healthcare fraud takedown involving 15 defendants charged, eight arrests, and more than $50 million in intended losses. This case study shows how SentinelGraph surfaced a related high-risk network through public-record linkage, entity resolution, and explainable risk scoring.

Important: This case study references public allegations announced by the U.S. Department of Justice on April 2, 2026. Charges are allegations, and defendants are presumed innocent unless and until proven guilty in court. SentinelGraph is an investigative prioritization tool, not an adjudication system.

The Public Enforcement Event

Federal prosecutors announced charges tied to more than $50 million in intended losses:

15

Defendants Charged

8

Defendants Arrested

$50M+

Intended Losses

  • •DOJ announced the matter on April 2, 2026, dubbed “Operation Never Say Die.”
  • •Public allegations included sham hospice operations and Medicare billing for beneficiaries who allegedly were not terminally ill.
  • •The Vance anti-fraud task force simultaneously suspended 221 California hospice and healthcare providers.
  • •One charged physician’s single NPI appeared across 63 different hospice agencies on over 6,000 claims totaling $35.8 million.

What SentinelGraph Saw

SentinelGraph elevated a network, not just a single entity. The system linked related organizations, people, and operational indicators into a single reviewable cluster, then assigned a high investigative priority score based on the concentration and interaction of risk signals.

Illustrative High-Signal Factors

Shared Infrastructure Clustering

Shared addresses, overlapping entity registrations, and phone number reuse across multiple hospice-related entities, suggestive of coordinated control.

Operational Network Pattern

Operational links among hospice-related entities that would be unusual for legitimate peer groups, including a single physician associated with dozens of agencies.

Repeated Control Relationships

Repeated individuals, officers, managers, or associated parties appearing across multiple filings and business records, including individuals already on federal exclusion lists.

Network-Level Operating Structure

Patterns indicating a coordinated network rather than isolated outliers, with entity creation dates, geographic clustering, and billing patterns collectively exceeding peer baselines.

Pre-Enforcement Signal Strength

Source-backed relationships significant enough to justify human review even before any public enforcement announcement.

Scoring disclaimer: Risk scores reflect investigative priority based on available inputs. They are not proof of misconduct or legal liability.

From First Signal to Source-Backed Escalation Package

For this network, SentinelGraph generated four report types that move an investigator from initial alert through deeper review and source validation.

Tier 1

Lead Brief

Initial high-risk summary designed for fast triage.

The Lead Brief gives the one-minute answer. It tells a reviewer who or what was flagged, how high the score was, why the network rose to the top, and whether the lead warrants immediate follow-up. Consumable by an executive, analyst, or counsel without a long read.

What a buyer should notice

  • • The score is clear and easy to interpret
  • • Top escalation reasons are concise and actionable
  • • The output looks like a triage product, not a black-box alert
Tier 2

Investigation Memo

Analyst-ready narrative that structures the case for deeper review.

The Investigation Memo turns a high-risk signal into a coherent narrative. It lays out the relationships that matter, explains why the cluster looks unusual, identifies the principal risks, and points the reviewer toward likely next questions.

Why this matters commercially

This is where SentinelGraph starts to feel operational. It does not just alert — it helps structure the analyst’s reasoning and reduce the time needed to get from signal to a meaningful review posture.

Tier 3

Full Dossier

Deep entity and network intelligence for serious review work.

The Full Dossier brings together the broader web of entities, addresses, roles, historical context, and relationship pathways needed to understand a network at investigation depth. It includes entity network tables, risk signal analysis, assertion chains, timeline, and recommended actions.

Presentation note

Public-facing versions of the Full Dossier are typically redacted. This report contains the richest expression of product value and analytical workflow.

Source Layer

Verified Source Summary

Source trace and validation layer that grounds the output in records.

Fraud intelligence products live or die on credibility. The Verified Source Summary shows that SentinelGraph findings can be traced to underlying records and source material rather than unsupported model output.

What a strong preview shows

  • • Source table with source type, date, and relevance
  • • Mapping between key findings and supporting records
  • • Clean distinction between source-backed facts and analytical interpretation
  • • A visibly disciplined, trust-building format

Why This Matters

This case illustrates SentinelGraph’s value not as a replacement for investigators, but as a force multiplier that helps teams focus scarce attention on the networks most likely to warrant deeper review.

For Investigators

Structured leads that surface network-level patterns, not just individual red flags, reducing the time from signal to meaningful review.

For Counsel

Evidence-traceable reports with source attribution, confidence framing, and clear separation between facts and analytical interpretation.

For Payors & Plans

Proactive identification of high-risk provider networks before enforcement actions, enabling earlier intervention and exposure reduction.

For Compliance Teams

Continuous monitoring of exclusion-linked control risk across provider networks, with escalation-ready reporting at multiple tiers.

Methodology and Guardrails

SentinelGraph uses structured data linkage, network analysis, and explainable risk scoring to elevate clusters for human review. Its findings are intended for investigative prioritization and must be assessed by qualified professionals using underlying source material and applicable legal standards.

Important Limitations

  • ●SentinelGraph does not determine guilt, innocence, or legal liability.
  • ●Alignment with later public enforcement is retrospective validation, not proof of causation.
  • ●Scores reflect relative investigative priority based on the data available at the time.
  • ●Findings depend on source quality, source coverage, and timing.
  • ●Human review remains essential.

Interactive Preview

See the report generation workflow

Click below to experience the same report generation flow subscribers see in their dashboard. Uses a fictional subject — Sentinel Grid Healthcare Holdings LLC — for demonstration.

Preview the report generation workflow subscribers experience

See how SentinelGraph can work for your team

Request a live demo, review a redacted sample report pack, or schedule a validation walkthrough using a known public case.

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.