Technical Specifications

Architecture Blueprint

A cloud-native, microservices-based platform deployed on FedRAMP-authorized infrastructure. IRIP operates across five integrated technical layers to ingest, analyze, and secure rural health data.

FedRAMP AWS/Azure
FHIR R4 API
Apache Kafka Stream
XGBoost & LightGBM
Microsoft Sentinel SIEM

Layer 1

Statewide Data Ingestion & HIE Integration

The foundational layer responsible for securely collecting clinical data from all ARHTP-connected systems and routing it downstream.

FHIR R4 API Gateway

Primary interface for modern EHR systems (Epic, Meditech, Athenahealth) to submit encounter records and vitals.

ALOHR Bidirectional Sync

Direct integration with the Alabama One Health Record (ALOHR). Every IRIP encounter enriches the shared state record.

Legacy HL7 v2 Translation

A Mirth Connect-based engine translates older EHR formats to FHIR R4, ensuring no rural facility is left behind due to legacy tech.

Real-Time RPM Stream

Apache Kafka event streaming processes high-frequency device telemetry with a target latency of under 60 seconds.

Layer 2

Population Health Risk Stratification

The core analytical capability. A continuously updated ML model assigning risk scores to every rural Medicaid beneficiary.

Model Architecture

Ensemble of XGBoost and LightGBM models trained on 2018-2025 Alabama Medicaid claims and ADPH vital statistics.

Feature Engineering

Over 140 features analyzed, including chronic condition burden, pharmacy adherence, and geographic distance to care.

Bias Monitoring & Audits

Accuracy and fairness metrics disaggregated by race and geography. Annual external audit conducted by UAB School of Public Health.

Inference Pipeline

Weekly batch recalculations + real-time recalculations triggered within 4 hours of high-acuity events (ED visit, hospital admit).

Layer 3

Real-Time Provider Dashboard Network

A lightweight, React.js web interface accessible at every ARHTP-connected facility, delivering actionable intelligence.

Care Gap Alerts

Prioritized lists of high-risk patients ranked by actionable care gaps, enabling 1-click care coordination workflows.

Telehealth Tracking

Real-time tracking of consultation volume, wait times, and outcomes across the hub-and-spoke ARHTP network.

Technical Stack

Frontend built in React.js/TypeScript served via CDN for low-bandwidth areas. Backend API built in Python (FastAPI). All access strictly governed by HIPAA minimum necessary standards and MFA.

Layer 4

Cybersecurity & Data Governance

Protecting vulnerable rural networks through zero-trust architecture and shared regional SOC operations.

Platform Security

AES-256 encryption at rest, TLS 1.3 in transit. Strict Zero-Trust Network Architecture with semi-annual third-party penetration testing.

Shared SOC Services

Five regional IT hubs operating Microsoft Sentinel SIEM. Providing 24/7 threat detection and alert triage for connected facilities.

Incident Response SLAs

NIST-aligned playbook execution with aggressive targets: Tier 1 (critical) response under 1 hour; Tier 2 (high) under 4 hours.

Layer 5

Automated CMS Accountability Engine

The administrative backbone transforming raw program activity into structured CMS reports, securing future funding.

Automated KPI Collection

Maps 50+ KPIs across ARHTP's 11 initiatives to direct data sources, eliminating error-prone manual spreadsheet submissions.

CMS Report Generation

Auto-generates quarterly CMS performance reports with validated metrics, reducing preparation time from 4 weeks to just 3 days.

Outcome Attribution

Maintains a longitudinal record linking specific ARHTP interventions (like RPM or Telehealth) directly to prevented hospitalizations to mathematically prove ROI.

Review the Full Technical Implementation Plan

Read the complete data governance protocols, FedRAMP alignment strategy, and phased deployment roadmap in the official proposal.

Download Full Technical Proposal