​
Eka Care Ekascribe Typescript SDK Integration

This guide explains how to integrate the Eka Care Ekascribe Typescript SDK into your application.

​
Overview

The Eka Care Ekascribe SDK allows you to capture and process audio, generating structured medical documentation using Eka Care’s voice transcription API.

​
Prerequisites

Before getting started, ensure you have:
  • Node 14 or higher
  • npm or yarn for dependency management
  • Access and refresh tokens from Eka Care (optional for some methods)
  • Microphone access via browser permissions
  • Stable network connectivity

​
Installation

Install the SDK using npm or yarn: 
npm install ekascribe-ts-sdk
# or
yarn add ekascribe-ts-sdk

​
Usage

​
1. Initialize Ekascribe

Before using any other method, initialize the SDK with access and refresh tokens. 
initEkascribe({
  access_token: '<your_access_token>',
  refresh_token: '<your_refresh_token>',
});

​
2. Fetch configurations list

Get supported input languages, output formats, and consultation modes. 
getEkascribeConfig();

  • ​
    Response type:

{
  data?: {
    supported_languages: [
      { id: '1', name: 'EN' },
      { id: '2', name: 'HI' }
    ];
    supported_output_formats: [
      { id: 'eka-emr-template', name: 'Eka EMR Format' },
      { id: 'clinical-notes-template', name: 'Clinical Notes' }
    ];
    consultation_modes: [
      {
        id: 'consultation',
        name: 'Consultation',
        desc: 'Eka Scribe will listen to your conversation and create clinical notes'
      },
      {
        id: 'dictation',
        name: 'Dictation',
        desc: 'Dictate your notes to Eka Scribe and create clinical notes'
      }
    ];
    max_selection: {
      languages: 2;
      output_formats: 2;
      consultation_mode: 2;
    };
  };
  message?: string;
  code?: number;
}

​
3. Start recording

Start recording with user-selected options. 
await startVoiceRecording({
  mode: 'consultation',
  input_language: ['te', 'en'],
  output_format_template: [{ template_id: 'eka_emr_template' }],
  txn_id: 'abc-123',
});

  • ​
    Response type:

{
  error_code?: ERROR_CODE;
  status_code: number;
  message: string;
  business_id?: string;
  txn_id?: string;
};

​
4. Pause recording

Use the method to pause voice recording 
await pauseRecording();

  • ​
    Response type:

{
  error_code?: ERROR_CODE;
  status_code: number;
  message: string;
  is_paused?: boolean;
};

​
5. Resume recording

Use the method to resume voice recording 
await resumeRecording();

  • ​
    Response type:

{
  error_code?: ERROR_CODE;
  status_code: number;
  message: string;
  is_paused?: boolean;
};

​
6. End recording

Use the method to end voice recording 
await endRecording();

  • ​
    Response type:

{
  error_code?: ERROR_CODE;
  status_code: number;
  message: string;
  failed_files?: ['1.mp3', '2.mp3']; // if there are any failed files
  total_audio_files?: ['1.mp3', '2.mp3', '3.mp3', '4.mp3']; // list of all audio files generated
};

​
7. Retry upload recording

Use this method to retry uploading failed audio files. 
await retryUploadRecording({ force_commit: true / false });

  • ​
    Response type:

{
  error_code?: ERROR_CODE;
  status_code: number;
  message: string;
  failed_files?: ['1.mp3', '2.mp3'];
  total_audio_files?: ['1.mp3', '2.mp3', '3.mp3', '4.mp3'];
};
force_commit behavior — If force_commit is set to true, the SDK will call the commit API even if some audio files still fail to upload after retrying once. — If force_commit is set to false, the SDK will wait until all audio files are uploaded successfully before making the commit request.

​
8. Cancel recording session

Use the method to cancel a recording session. 
await cancelRecordingSession({ txn_id: 'abc-123' });

  • ​
    Response type:

{
  status: string;
  message: string;
  code: number;
  error?: {
    code: string;
    message: string;
    display_message: string;
  };
}

​
9. Commit transaction

Use this method to commit a transaction that is not yet committed or returned a “commit failed” error in a previous step. 
await commitTransactionCall();

  • ​
    Response type:

{
  error_code?: ERROR_CODE;
  status_code: number;
  message: string;
};

​
10. Stop transaction

Use this method to stop a transaction that has not yet been stopped or returned a “stop failed” error in a previous step. 
await stopTransactionCall();

  • ​
    Response type:

{
  error_code?: ERROR_CODE;
  status_code: number;
  message: string;
};

​
11. Get output template prescriptions

Use this method to fetch the final generated prescription output for a session. 
await getTemplateOutput({ txn_id: 'abc-123' });

​
12. Get total uploaded files

Use this method to retrieve all the audio files generated for a specific session. 
const files = await getTotalAudioFiles();

  • ​
    Response type:

['1.mp3', '2.mp3', '3.mp3', '4.mp3'];

​
13. Get successfully uploaded files

Use this method to retrieve all the audio files that were uploaded successfully. 
const successFiles = await getSuccessfullyUploadedFiles();

  • ​
    Response type:

['3.mp3', '4.mp3'];

​
14. Get failed audio files

Use this method to retrieve all the audio files that failed to upload. 
const failedFiles = await getFailedFiles();

  • ​
    Response type:

['1.mp3', '2.mp3'];

​
15. Generic Error Callback

Whenever an error occurs in the SDK during voice recording, the following callback will be triggered. You can listen to this to handle errors globally. 
onError(({ error_code, status_code, message }) => {
  console.error('Ekascribe SDK Error:', { error_code, status_code, message });
});

​
ERROR CODES:

Error CodeDescription
microphoneMicrophone access error (permission denied or unavailable)
txn_init_failedFailed to initialize transaction
txn_limit_exceededMaximum number of concurrent transactions exceeded
unknown_errorAn unknown or unclassified error occurred
txn_stop_failedError occurred while stopping the transaction
audio_upload_failedAudio file upload to server failed
txn_commit_failedCommit call failed for the current transaction
invalid_requestRequest to SDK was malformed or missing required parameters
vad_not_initializedVoice activity detection engine was not initialized
no_audio_captureNo audio was captured during the recording session
txn_status_mismatchInvalid operation due to mismatched transaction status