Infrastructure

Infrastructure you can deploy, monitor, and govern.

AcuityAI connects edge capture to HAPI FHIR normalization, OpenEMR workflows, real-time ML diagnostics and risk scoring, and dashboards — built for reliability and clinical operations.

Request a DemoDownload Architecture Overview

End-to-end architecture

Edge → FHIR → EHR → AI → Dashboards

Layer 1

Edge Layer

AcuityAI Clinic

Kiosk / device capture workflows

Structured intake: vitals, symptoms, diagnostic outputs

Quality checks (completeness, plausibility)

Layer 2

FHIR Canonical Layer

HAPI FHIR

FHIR R4 normalization (validated resources)

Terminology alignment

Event traceability and audit logging

Layer 3

Clinical Workflow Layer

OpenEMR

Encounter integration and chart updates

Tasks / alerts for escalation pathways

Structured note insertion / summary blocks

Layer 4

ML Scoring Layer

AI Engine

Real-time diagnostics + risk scoring services

Model versioning + traceability

Monitoring of inputs and scoring behavior

Layer 5

Dashboards

Operational Visibility

Operational KPIs (flow, utilization, bottlenecks)

Clinical oversight views (acuity mix, escalations)

Data quality / completeness monitoring

FHIR resource coverage

Canonical FHIR R4 resources.

All point-of-care data is normalized into validated FHIR R4 resources, enabling consistent downstream integration and reporting.

PatientEncounterObservationConditionDiagnosticReportServiceRequestProcedureMedicationRequestMedicationAdministration

Integration mechanisms

How the layers connect.

RESTful FHIR APIs

All data flows through standardized FHIR REST APIs to HAPI, enabling consistent downstream integration.

OpenEMR workflow mapping

Outputs map directly into OpenEMR modules and UI touchpoints — tasks, notes, summaries, alerts.

Scoring service invocation

ML scoring services are invoked at defined workflow points, such as post-capture, with full traceability.

Operational readiness

Built for day-2 operations.

Observability

Logs and metrics for capture, ingestion, scoring, and workflow delivery across the full pipeline.

Auditability

A complete trace of what happened, when, and why — from edge capture through to score output.

Reliability controls

Retries, validation gates, and safe failure modes ensure the pipeline stays stable under pressure.

Governance controls

Model versions and workflow configuration are fully traceable for clinical and IT governance.

Want to go deeper on the architecture?

We will walk through the full stack with your engineering and clinical informatics teams.

Talk to an Integration Lead