noun: vault; a secure store where valuables are kept safe.

Audio Vault

A web service which stores audio segments (WAV) from a Voice Recognition / Digital Dictation system.

Audio Vault's key features:

• Handles the safe storage of audio segments.
• Schedules the processing of the audio segments to normalise the Bit Rates.
• Concatenates the audio segments into a single audio recording for use by medical secretaries.
• Interfaces with the hospitals document management system to attach to the audio file to the letter.
• Handles the safe storage of audio orphans.
• Implements audio retention polices.

After individual segments are stitched together the system generates a wav form image of the resulting audio dictation.

Future versions will incorporate peaks.js to provide real-time wav form syncronisation with the audio playback.

Architectural Decisions

• The system will be developed using the Go programming language so that it can be cross compiled to Linux and Windows.
• The system will be run from a primary server (node) and will use sync tools to keep a warm standby server (node) in a ready state.
• The web service should be run via a reverse proxy from a web server, e.g. IIS.
• The web service can be installed on Windows as a service using the NSSM tool.
• The system will use an SQLite database to store its meta data.
• The system will use SQLite Rsync to backup the SQLite database to a warm standby server.
• The system will use rClone to sync audio to a warm standby server.
• The system will offload audio processing to SoX Sound eXchange.
• The system will offload audio wav form generation to audiowaveform

Endpoints

A list of endpoints that the service will expose. The table outlines which endpoints use logging and which require an API key (configured in the settings.ini file)

endpoint	API Key	description
/health-check	N	Used by clients and automated checks to see if the service is running
/orphan	Y	Allows clients to submit audio not associated with a specific letter
/record	Y	Allows clients to submit meta data about incoming audio segments
/server-side-events	N	Sever push of CPU usage and list of pending segments that are being processed
/store	Y	Allows clients to submit audio segments for safe keeping
/stream	N	Allows web user interface to stream audio to a HTML5 audio element
/testing	N	A web user interface to allow tester to monitor the audio conversion process

Running Web Service from IIS using a Reverse Proxy

To configure IIS with Reverse Proxy Request Routing support, you'll need to install application-request-routing

Then set-up a basic web.config file with a single rewrite rule to forward all requests/traffic to the web service. Substitute the port number, if you've changed it from the default of 1969.

<?xml version="1.0" encoding="UTF-8"?>
<configuration>
    <system.webServer>
    <rewrite>
        <rules>
            <rule name="ReverseProxyAudioVault" enabled="true" stopProcessing="true">
                <match url="(.*)" ignoreCase="true" />
                <action type="Rewrite" url="http://localhost:1969/{R:1}" />
            </rule>
        </rules>
    </rewrite>
    <httpProtocol>
        <customHeaders>
            <remove name="X-Powered-By" />
        </customHeaders>
    </httpProtocol>
    </system.webServer>
</configuration>