ShedBoxAI for Healthcare

data_sources:
  ehr_export:
    type: csv
    path: "patient_records.csv"

  lab_results:
    type: csv
    path: "lab_results.csv"

processing:
  relationship_highlighting:
    ehr_export:
      link_fields:
        - source: "ehr_export"
          source_field: "patient_id"
          to: "lab_results"
          target_field: "patient_id"

  contextual_filtering:
    ehr_export:
      - field: "last_visit"
        condition: ">2024-01-01"
        new_name: "recent_patients"

output:
  type: file
  path: "patient_summary.json"
  format: json

data_sources:
  appointments:
    type: csv
    path: "appointments.csv"

processing:
  advanced_operations:
    dept_metrics:
      source: "appointments"
      group_by: "department"
      aggregate:
        total_appointments: "COUNT(*)"
        avg_wait_time: "AVG(wait_minutes)"
        total_no_shows: "SUM(no_show)"
      sort: "-total_appointments"

output:
  type: file
  path: "operations_metrics.json"
  format: json

data_sources:
  clinical_notes:
    type: csv
    path: "notes_export.csv"

ai_interface:
  model:
    type: rest
    url: "https://api.anthropic.com/v1/messages"
    method: POST
    headers:
      x-api-key: "${ANTHROPIC_API_KEY}"
      Content-Type: "application/json"
    options:
      model: "claude-sonnet-4-20250514"

  prompts:
    summarize:
      system: "You are a medical documentation assistant."
      user_template: |
        Summarize these clinical notes concisely:

        {% for note in clinical_notes %}
        Patient {{ note.patient_id }}: {{ note.note_text }}
        {% endfor %}

        Provide key findings and action items.

output:
  type: file
  path: "note_summaries.md"
  format: json