Medix Prior Auth — Command Center

Total Agents

Active in pipeline

Extraction

Reader + Core agents

Policy & Matching

Matcher + Payer agents

Per-Drug Agents

Drug + Payer specialists

Docs Processed

Total referrals

Appeal & Evidence

Appeal + P2P + Evidence

Validation

Quality + Cross-check

Accuracy

Audit + Regression

Oversight

Arch + Security + Accuracy

Watchdog

Problem monitors

Problem Watchdog Map

Loading watchdog data...

Reader Agents — Document Extraction

Core Pipeline Agents

Policy Matcher Agents — 4-Layer Verification

Payer Policy Agents

Per-Drug Agents (BCBSTX Phase 1)

Interpretation & Validation Agents

Appeal, Evidence & Medical Necessity Agents

Prediction, Workflow & Generation Agents

Lifecycle, Eligibility & Urgency Agents

Accuracy Enforcement Agents — 99% Target

Oversight Agents — Architecture, Security, Accuracy

Analytics & Monitoring Agents

Denial Prevention Agents

Watchdog Agents — Problem Prevention

Denial Tracker Agents (Per-Drug Per-Payer)

Pharma Data Intelligence Agents

V2 System Architecture

  REFERRAL PDF
       |
       v
  [Document Loader] -------- PDF to Text + OCR Fallback + Redaction Detection
       |
  [Clinical Interpretation] - Abbreviation expansion, OCR correction, date normalization
       |
       |  +----- PARALLEL EXTRACTION (8 Reader Agents) -----+
       |  |                                                   |
       v  v                                                   |
  [Admin/Eligibility Reader]   [Symptoms Reader]              |
  [Medical History Reader]     [Medications Reader]           |
  [Prior Testing Reader]       [Labs Reader]                  |
  [Vitals/Exam Reader]         [Referral Order Reader]        |
       |  |                                                   |
       |  +---------------------------------------------------+
       |                    |
       |         [Extraction Audit Agent] -- validates ranges, dates, formats
       |         [Cross-Validation Agent] -- reader contradictions
       |         [Quality Gate Agent] ------ documentation completeness
       v
  [Drug Classifier] ----------- Classifies drug and indication (Claude API)
       |
       v
  [Eligibility Verification] -- member ID format, plan type, PCP referral
  [Urgent Detection] ---------- 4 urgency levels, 13 clinical triggers
       |
       v
  +---- 4-LAYER POLICY MATCHING (5 Payer Agents) ----------+
  |  BCBSTX | Aetna | UHC | Cigna | Medicare               |
  |                                                          |
  | [Layer 1: Eligibility] ---- admin, referral, signed order|
  | [Layer 2: Clinical Doc] --- symptoms, vitals, labs       |
  | [Layer 3: Medical Necessity] -- ICD-10, step therapy, dx |
  | [Layer 4: Drug-Specific] ------- Step therapy, biosimilar, labs |
  |                                                          |
  +----------------------------------------------------------+
       |                    |
       |         [Matcher Audit Agent] ----- false positive/negative detection
       |         [Policy Validation Agent] -- policy JSON integrity
       v
  [Readiness Aggregator] ------ GREEN / YELLOW / RED
  [Denial Prediction Agent] --- risk scoring, denial triggers
  [Next Steps Agent] ---------- prioritized action items
       |
       v
  AUTHORIZATION PACKAGE -------- Ready for payer submission
       |
       v  (if denied)
  [Appeal Letter Generator] --- clinical evidence + guideline citations
  [P2P Call Prep Agent] ------- physician talking points
  [Clinical Evidence Agent] --- landmark trials, society guidelines, LCD precedent
  [Medical Necessity Agent] --- patient-specific necessity arguments
  [Auth Status Tracker] ------- 18-state lifecycle tracking
       |
       v
  [Patient Communication] ----- plain-language letters (6th-grade level)

  OVERSIGHT:   [Architecture Review] [Security Review] [Accuracy Validation]
  ACCURACY:    [Extraction Audit] [Matcher Audit] [Regression Gate] [Policy Validation]
  WATCHDOGS:   [OCR Quality] [Reader Consistency] [Policy Match] [Drugs] [Redaction]
  ANALYTICS:   [Denial Analytics] [Policy Monitor] [Industry ROI]

Operational Savings

Staff time + denial prevention + clawback avoidance

Total Revenue Impact

Including patient retention + chair utilization + brand documentation

Hours Saved

Based on documents processed

Patients Retained

Who would have abandoned treatment due to PA delays

Denials Prevented

Pre-submission checks caught missing items

Operational Savings

Staff Time Savings

Auth coordinators average 45 min/PA manually. AI processes in under 1 minute.

Denial Prevention

Pre-submission checks catch step therapy gaps, missing TB screening, coding errors. Each denied claim costs $118 to rework.

Clawback Prevention

Infusion drugs average $5,200 per clawback. Proper PA documentation prevents post-payment recoupment.

Rework Reduction

22% of PAs require resubmission. Completeness checks reduce rework by 65%.

P2P Call Prep

AI generates P2P prep with drug-specific talking points in 1 min vs 30+ min manual preparation.

Appeal Letter Prep

AI drafts appeals with ACR/AAN/ACG guideline citations and drug-specific evidence in 2 min vs 45 min.

Infusion Center Revenue Impact

Patient Retention

18% of patients abandon treatment due to PA delays. Each patient = ~$85K/year in infusion revenue. Faster PA keeps patients in chairs.

Chair Utilization Recaptured

Faster PA approval means earlier scheduling. Every day of PA delay = $450 in lost chair revenue per patient.

Brand Drug Documentation

When medically justified (biosimilar failure, adverse reaction), proper documentation supports brand approval. Remicade vs biosimilar = $5,700/infusion differential.

Per-Drug Revenue Analysis

Drug	J-Code	Processed	$/Infusion	Revenue Processed	Denials Prevented	Revenue Protected
Loading drug data...

Per-Document Value

Minutes Saved Per PA

vs. industry average of 45 minutes manual processing

Dollar Value Per PA

Operational savings per document processed

Industry Benchmarks vs. Medix Prior Auth

Time & Efficiency

PA Processing Time 45 min → <1 min

Appeal Letter Drafting 45 min → 2 min

P2P Call Preparation 30 min → 1 min

Days to PA Approval 14 days → ~3 days

Staff FTEs per 500 PAs/mo 2.5 FTE → 0.5 FTE

Infusion Center Revenue Impact

Initial Denial Rate 24% → ~14%

Patient Abandonment (PA delays) 18% → ~9%

PA Rework Rate 22% → ~8%

Clawback Rate 4.0% → ~1.6%

Avg Clawback Amount $5,200

Remicade vs Biosimilar Premium $5,700/infusion

Annual Impact Projections

Monthly PAs	Hours Saved	Operational Savings	Retention Impact	Chair Utilization	Total Annual Impact
Loading projections...

Data Sources: AMA Prior Authorization Physician Survey (2023), CAQH Index (2023), MGMA DataDive Cost and Revenue, HFMA Revenue Cycle Benchmarking Report, Experian Health State of Claims Report, Specialty Infusion Industry Benchmark Report (2024). Drug reimbursement based on ASP+6% Medicare Part B rates. Patient abandonment data from AMA specialty infusion PA impact study. Staff rates from Bureau of Labor Statistics (2024).

Total in Pipeline

Active cases

Need Appeal Letter

Awaiting generation

Need P2P Prep

Calls to prepare

Awaiting Decision

Submitted / in review

Available Agents

Ready to assist

Denial & Appeal Cases

Loading appeal cases...

Agents Helping

Loading agents...

Specialties Tracked

Medical specialties

Active

With live drug agents

Planned

Upcoming phases

Drugs Active

Per-drug agents live

Drugs Planned

In expansion pipeline

Specialty Intelligence Agents

Loading specialty data...

Cases Analyzed

Total referrals processed

Unique Drugs

Distinct drugs tracked

Data Categories

Intelligence streams

HIPAA Compliant

De-identified & Aggregate

Minimum N=10 for reporting

Drug-Level Intelligence Summary

Loading drug data...

Data Categories for Pharma Partners

Loading data categories...

Pharma Data Monetization: All data is HIPAA-compliant, de-identified, and reported in aggregate only (minimum N=10). Data streams include prescribing patterns, real-world outcomes (RWE), market access intelligence, biosimilar adoption dynamics, and competitive intelligence. Updated weekly/monthly/quarterly depending on category. Valuable to pharmaceutical companies for market research, post-market surveillance, payer negotiations, and competitive strategy.