HL7 v2 → FHIR pipeline
Inbound MLLP listener, Matchbox $transform with StructureMaps, HAPI FHIR write. Durable message store with retries and a DLQ for anything that fails.
If you need to be able to interface your existing EHR with apps that require FHIR subscriptions, now you can do it without having to change what you already have. Your EHR talks to our service and our service talks FHIR to the subscribers. We can interface with your EHR using standard HL7 messages it already speaks or in some cases, using the EHRs own proprietary apis.
You can run your own subscription service to interface your EHR to applications using the standard FHIR subscriptions either on your own prem and hardware, or in the cloud, or you can have us manage the service for you.

Captured from subscription-service-ui.bzonfhir.com. The whole site is one click away.
Inbound MLLP listener, Matchbox $transform with StructureMaps, HAPI FHIR write. Durable message store with retries and a DLQ for anything that fails.
Vendor-specific REST APIs (Athena, others) polled on a schedule and normalized into the same FHIR R4 surface — no HL7 v2 required.
Standards-compliant Subscriptions Backport. Downstream apps register criteria; subscription-service fires notifications when matching resources land.
Eight purpose-built routes for the ops humans actually running the system. Dashboard, interfaces, messages, subscriptions, matchbox, DLQ, audit, settings.
JSON logs, Prometheus metrics, OpenTelemetry traces, FHIR AuditEvent persistence, and an agent-queryable admin API. The raw signals, no paywall.
Drop-in YAML + StructureMap bundles for Epic, Meditech, Cerner, athena, NextGen. Profiles handle the quirks; your pipeline stays standards-first.
Bring the whole stack up locally with one docker compose
command. No license server to call. No registration form. If you can
run Docker, you can run subscription-service.
# 1. Clone the repo
git clone [email protected]:bzimbelman/fhir-ehr-subscriptions-service.git
cd fhir-ehr-subscriptions-service/deploy/docker
# 2. Bring HAPI + Matchbox + interface-engine + operator UI up
docker compose up -d A working demo is worth ten pages of feature copy. The operator UI is live, no signup, no demo-data slideshow — it's the real app pointed at a sandbox FHIR server.