Open Compute

Open-source agentic AI for health-tech

Open Compute is a powerful Python library that uses AI agents to transform patient journeys into standards-compliant FHIR (Fast Healthcare Interoperability Resources) data. Built by Jori Health, this tool enables healthcare developers to bridge the gap between narrative patient experiences and structured healthcare data.

Quick Start

Installation

pip install open-compute

Requirements

• Python 3.9+
• OpenAI API Key: Set as environment variable

  export OPENAI_API_KEY='your-api-key-here'
  

Usage

Patient Journey to FHIR Conversion

The primary use case is converting narrative patient journeys into structured, validated FHIR resources:

from open_compute import (
    PatientJourney,
    JourneyStage,
    generate_fhir_from_journey,
)

# 1. Define a patient journey with clinical narrative
journey = PatientJourney(
    patient_id="patient-123",
    summary="58 year old male presents to ER with chest pain, diagnosed with acute MI",
    stages=[
        JourneyStage(
            name="Registration",
            description="Patient registered in ER",
            metadata={
                "timestamp": "2024-01-15T10:30:00Z",
                "location": "Emergency Department"
            }
        ),
        JourneyStage(
            name="Triage",
            description="Initial assessment - chest pain, elevated BP",
            metadata={
                "vital_signs": {
                    "blood_pressure": "150/95 mmHg",
                    "heart_rate": "88 bpm"
                },
                "chief_complaint": "Chest pain"
            }
        ),
        JourneyStage(
            name="Diagnosis",
            description="Diagnosed with acute myocardial infarction",
            metadata={
                "condition": "Acute MI",
                "icd10_code": "I21.9"
            }
        ),
        JourneyStage(
            name="Treatment",
            description="Administered aspirin and nitroglycerin",
            metadata={
                "medications": ["Aspirin 325mg", "Nitroglycerin 0.4mg"]
            }
        ),
    ]
)

# 2. Generate FHIR resources using AI
result = generate_fhir_from_journey(
    journey=journey,
    patient_context="58 year old male named John Doe with history of hypertension",
    model="gpt-4o-mini",  # or "gpt-4", "gpt-3.5-turbo"
    fhir_version="R4",
    max_iterations=3,
    auto_save=True  # Saves to output/john_doe/
)

# 3. Check results
print(f"✅ Success: {result.success}")
print(f"📊 Generated {len(result.generated_resources)} FHIR resources")
print(f"🔄 Iterations: {result.iterations}")

# View generated resource types
for resource in result.generated_resources:
    print(f"  - {resource['resourceType']}/{resource.get('id', 'no-id')}")

Output Structure

When auto_save=True, generated files are saved to output/{firstname_lastname}/:

output/john_doe/
├── patient_bundle.json    # Complete FHIR Bundle (all resources)
├── bulk_fhir.jsonl       # Bulk FHIR format (one resource per line)
└── README.txt            # Summary of generated resources

Advanced Configuration

from open_compute import AIJourneyToFHIR

# Create agent with custom configuration
agent = AIJourneyToFHIR(
    api_key="your-openai-key",  # or use env var OPENAI_API_KEY
    model="gpt-4o-mini",
    fhir_version="R4",
    max_iterations=5,
    max_fix_retries=3,
    auto_save=True,
    save_directory="output",
    parallel_generation=True,  # Faster generation
    use_enhanced_context=True  # Better accuracy with FHIR profiles
)

# Generate resources
result = agent.generate(journey, patient_context="...")

Examples

We provide comprehensive examples to help you get started:

Running the Examples

# Make sure you have your OpenAI API key set
export OPENAI_API_KEY='your-api-key-here'

# Run the main example
python examples/patient_journey_to_fhir_example.py

Available Examples

The example demonstrates:

• Creating a patient journey with multiple stages
• Generating FHIR resources (Patient, Encounter, Observation, Condition, MedicationStatement, Procedure)
• Validating generated resources
• Auto-saving to organized output directory

Contributing

We welcome contributions! Open Compute is an open-source project and we'd love your help.

Areas for Contribution

• Additional FHIR resource types
• Support for more FHIR versions
• Enhanced validation rules
• Performance optimizations
• Documentation improvements

Getting Started

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Make your changes
4. Run tests (pytest)
5. Commit your changes (git commit -m 'Add amazing feature')
6. Push to the branch (git push origin feature/amazing-feature)
7. Open a Pull Request

License

MIT License - see LICENSE for details

Made with 🏥 for better healthcare interoperability