Skip to main content
This document provides a guide on how to upload audio files using a presigned URL of previous step.

​
Prerequisites

Before uploading files, you must first obtain a presigned URL by calling the Get Presigned URL endpoint.

​
Upload Process

​
Step 1: Use the Presigned URL API Response

Use the response from the presigned URL API call. The response will look like this: 
upload_response = {
    "uploadData": {
        "url": "https://m-prod-ekascribe-batch.s3.amazonaws.com/",
        "fields": {
            "x-amz-meta-mode": "dictation",
            "x-amz-meta-uhid": "unique_patient_id",
            "x-amz-meta-hfid": "unique_health_facility_id",
            "key": "EC_173210496011417/txn_301/20250617_105524/${filename}",
            "x-amz-algorithm": "AWS4-HMAC-SHA256",
            "x-amz-credential": "...",
            "x-amz-date": "...",
            "policy": "...",
            "x-amz-signature": "..."
        }
    },
    "folderPath": "EC_173210496011417/txn_301/20250617_105524/",
    "txn_id": "txn_301"
}

​
Step 2: Add Your Audio File Paths

Specify the paths to your audio files that you want to upload: 
# Add your audio file paths here
audio_files = [
    "/path/to/your/first_audio.wav",
    "/path/to/your/second_audio.mp3",
    "/path/to/your/third_audio.m4a"
]

​
Step 3: Set the Correct Action

For EkaScribe V2, use the following action: 
action = "ekascribe-v2"

​
Complete Upload Workflow

Here’s the complete Python code combining all three steps: 
import requests
import os

def upload_file(upload_data, folder_path, file_path):
    """Upload a single audio file"""
    file_name = os.path.basename(file_path)
    upload_data['fields']['key'] = folder_path + file_name
    
    with open(file_path, 'rb') as file:
        files = {'file': (file_name, file)}
        response = requests.post(upload_data['url'], data=upload_data['fields'], files=files)
    
    if response.status_code == 204:
        return {'key': folder_path + file_name, 'size': os.path.getsize(file_path)}
    else:
        raise Exception(f"Upload failed: {response.status_code}")

# Step 1: Use the presigned URL response (from previous API call)
upload_response = {
    "uploadData": {
        "url": "https://m-prod-ekascribe-batch.s3.amazonaws.com/",
        "fields": {
            "x-amz-meta-mode": "dictation",
            "x-amz-meta-uhid": "unique_patient_id",
            "x-amz-meta-hfid": "unique_health_facility_id",
            "key": "EC_173210496011417/txn_301/20250617_105524/${filename}",
            "x-amz-algorithm": "AWS4-HMAC-SHA256",
            "x-amz-credential": "...",
            "x-amz-date": "...",
            "policy": "...",
            "x-amz-signature": "..."
        }
    },
    "folderPath": "EC_173210496011417/txn_301/20250617_105524/",
    "txn_id": "txn_301"
}

# Step 2: Add your audio file paths
audio_files = [
    "/path/to/your/first_audio.wav",
    "/path/to/your/second_audio.mp3",
    "/path/to/your/third_audio.m4a"
]

# Step 3: Set the correct action
action = "ekascribe-v2"

# Get transaction ID from upload response
transaction_id = upload_response["txn_id"]

# Upload audio files
upload_results = []
for audio_file in audio_files:
    result = upload_file(
        upload_response['uploadData'].copy(),  # Use copy to avoid modifying original
        upload_response['folderPath'],
        audio_file
    )
    upload_results.append(result)
    print(f"Uploaded: {result['key']}")

# Process responses
for response in upload_results:
    print(f"File: {response['key']}, Size: {response['size']} bytes")

​
Expected Response

After successful upload, you’ll see output like: 
Uploaded: EC_173210496011417/txn_301/20250617_105524/first_audio.wav
File: EC_173210496011417/txn_301/20250617_105524/first_audio.wav, Size: 2847392 bytes
Uploaded: EC_173210496011417/txn_301/20250617_105524/second_audio.mp3
File: EC_173210496011417/txn_301/20250617_105524/second_audio.mp3, Size: 1823456 bytes

​
Next Steps

After successfully uploading your audio files, proceed to Initialize Transaction to start the transcription process. Keep in mind that the same txn_id should be used while initializing transaction and fetching the output.