Skip to main content
fi-fhir docs

fi-fhir User Guide

Welcome to fi-fhir, a format-agnostic healthcare integration platform that transforms legacy formats (HL7v2, CSV, EDI X12, CDA/CCDA) into semantic events and routes them through configurable workflows.

Table of Contents

  1. Getting Started
  2. Core Concepts
  3. CLI Reference
  4. Source Profiles
  5. Workflow Configuration
  6. Terminology Management
  7. FHIR Output
  8. LLM-Powered Features
  9. Playground Tutorial
  10. Browser Playground
  11. Mapping Studio (IDE)
  12. Connections
  13. End-to-end demo
  14. Verification
I want to...Read...
Parse my first HL7v2 messageGetting Started
Understand the 3-phase pipelineCore Concepts
Configure a source profileSource Profiles
Route events to multiple destinationsWorkflow Configuration
Test workflows without side effectsWorkflow Configuration
Load standard terminologies (LOINC, RxNorm)Terminology Management
Map local codes to standard terminologiesTerminology Management
Generate FHIR resourcesFHIR Output
Get AI-powered warning explanationsLLM-Powered Features
Generate workflows from natural languageLLM-Powered Features
Search terminology by meaningLLM-Powered Features
Try it in the browserPlayground Tutorial
Know what the browser playground runs, and what needs the IDEBrowser Playground
Find my way around the IDEMapping Studio (IDE)
Click through the real IDE without installing anythingHosted demo: the IDE shell with a preview-only identity and no database; HL7 Preview works, every other surface shows its honest "not available on this deployment" state, nothing is stored
Define a source or destination connectionConnections
See what this replica is actually runningConnections
Pull real messages from a source into a sessionConnections
Run the batch-to-hospital demo on the LAN environmentEnd-to-end demo
Check what the engine admitted and deliveredVerification

What is fi-fhir?

fi-fhir addresses a core problem in healthcare integration: users think in workflow terms, but tools require format-specific knowledge.

Instead of writing code that references PID.3.1 or OBX.5, you work with semantic events like patient_admit and lab_result. The library handles format parsing, field mapping, validation, and routing automatically.

Key Features

  • Multi-format parsing: HL7v2, CSV/flatfiles, EDI X12, CDA/CCDA
  • Profile-driven: Each data feed gets its own configuration
  • Semantic events: Work with business concepts, not format details
  • Workflow DSL: YAML-based routing with CEL expression filters
  • FHIR R4 output: US Core profile mapper with 24+ resources
  • Production-ready: Retry, circuit breaker, observability built-in

Architecture Overview

Input Formats          Semantic Layer           Output/Actions
─────────────         ────────────────         ──────────────
HL7v2    ──┐          ┌─────────────┐          ┌─> FHIR API
Flatfile ──┼──────────┤ Canonical   ├──────────┼─> REST Webhook
EDI X12  ──┤          │ Event Model │          ├─> Database
CDA/CCDA ──┤          └─────────────┘          └─> Message Queue
FHIR     ──┘

Prerequisites

  • For CLI usage: Go 1.21+ or Docker
  • For TypeScript SDK: Node.js 18+
  • For Playground: Just a web browser!

Next Steps

If you're new to fi-fhir, start with Getting Started for a hands-on tutorial.

If you prefer to learn by exploring, try the Playground Tutorial to experiment in your browser.

See Also