Sahl AI Note Generation API
Overview
The Note Generation API takes a list of transcript items and produces a structured clinical note. This API is designed to streamline the process of creating comprehensive medical notes from transcribed encounters.
Endpoint
URL: https://sau-api.sahl.ai/note/generate
Method: POST
Authentication
Include your bearer token in the Authorization header:
Authorization: Bearer <YOUR_TOKEN>
Request Body
Content-Type: application/json
{
"output_objects": ["note"],
"language": "en-US",
"transcript_items": [
{
"text": "Patient reports persistent headaches.",
"speaker": "doctor"
}
],
"section_style": "paragraphs",
"split_by_problem": false,
"note_template": "GENERAL_MEDICINE",
"note": {
"sections": [
{
"key": "CHIEF_COMPLAINT",
"title": "Chief Complaint",
"text": "Persistent headaches"
}
]
},
"patient_context": "John Doe, 45-year-old male, history of migraines"
}
Parameters
output_objects
(array of strings, required): Specify which items to include in the response. Currently, only "note" is supported.language
(string, required): Language of the transcript and desired note. Options: "en".transcript_items
(array of objects, required): List of transcript items from the encounter.note_template
(string, optional): Template for the generated note. Options include "GENERAL_MEDICINE", "SOAP", etc.
Response
Success Response (200 OK)
Content-Type: application/json
[
{
"title": "Chief Complaint",
"body": [
"- 30-year-old patient experiencing shortness of breath for two days, especially when climbing stairs or exerting physical effort."
]
},
{
"title": "History of Presentation",
"body": [
"- Shortness of breath began suddenly two days ago.",
"- Symptoms occur with exertion, such as climbing stairs or walking a few meters.",
"- Patient reports feeling tired but denies fever, cough, headache, visual disturbances, joint pain, and swelling in the lower limbs.",
"- Able to walk to the nearby mosque, indicating some preserved functional capacity."
]
},
{
"title": "Past medical and surgical history",
"body": [
"- Patient has a history of hypertension diagnosed since childhood.",
"- Congenital heart disease with weakened heart muscle."
]
},
{
"title": "Allergy",
"body": ["- No allergy history discussed."]
},
{
"title": "Medication",
"body": [
"- Patient is currently taking antihypertensive medications regularly."
]
},
{
"title": "Social history",
"body": [
"- No information provided regarding smoking habits, alcohol consumption, lifestyle, social circumstances, functional status, or occupation."
]
},
{
"title": "Family history",
"body": ["- No family history of diseases discussed."]
},
{
"title": "Examination",
"body": [
"- Patient appears tired but oriented to time, place, and space.",
"- On auscultation, normal heart sounds S1 and S2 with a pan-systolic murmur were noted.",
"- Clear air entry in the lungs with some basal crackles.",
"- No lymphedema observed."
]
},
{
"title": "Diagnosis/Impression",
"body": [
"- Shortness of breath likely related to underlying congenital heart disease and hypertension.",
"- Possible heart failure needs to be evaluated by a cardiologist."
]
},
{
"title": "Plan",
"body": [
"- Patient to continue antihypertensive medications.",
"- ECG and further examination by a cardiac specialist are planned.",
"- Urgent referral to a cardiologist expected within the week due to the severity of symptoms.",
"- Advised to avoid excessive cold exposure."
]
}
]
Error Response (400 Bad Request)
{
"error": "Invalid request",
"message": "Detailed error message"
}
Notes
- The API generates a structured clinical note based on the provided transcript items and additional context.
Example Usage
curl -X POST 'https://sau-api.sahl.ai/note/generate' \
-H 'Content-Type: application/json' \
-H 'Authorization: Bearer <YOUR_TOKEN>' \
-d '{
"output_objects": ["note"],
"language": "en-US",
"transcript_items": [
{
"text": "Patient reports persistent headaches.",
"speaker": "doctor"
}
],
"section_style": "paragraphs",
"note_template": "GENERAL_MEDICINE"
}'
This API allows for efficient generation of structured clinical notes from transcribed medical encounters, enhancing the documentation process for healthcare providers.