Sentinel Setup Guide: Preparing EvoVoice for AI-Powered Insights

This guide provides step-by-step instructions to configure EvoVoice for AI-powered analysis. Before EvoVoice can process calls for AI insights, it requires a recording of each call. The recording is processed by AI, and prompts are applied to generate efficient outputs for the Sentinel process. This document explains how to prepare audio recordings for processing and how to set up the system to provide insights in the dashboard.

Step 1: Preparing the HostedSuite Insights Flow

The first step is to create a HostedSuite Insights Flow. This will act as a subflow, allowing multiple workflows to share it for AI call analysis.

Create a New Flow

1. Open the Flow Designer.

2. Click Create New Flow and enter a name.

   • Example: "HostedSuite AI Call Insights"

Add a Recording File Parameter

To ensure each call recording is properly processed, we must define a Recording File parameter. This will receive the audio file before it is analyzed.

Steps to Add a Parameter

1. In the Flow Designer, select the HostedSuite Insights flow.

2. Click the Parameters tab.

3. Click Add Parameter, then fill in the details:

- **Name**: `RecordingFile`

- **Type**: `AudioFile`

- **Public**: ✅ *(Check this option)*

4. Click Add Parameter.

Once added, the new parameter should appear in your HostedSuite AI Call Insights node.

Step 2: Configuring the Node Fields

Now that we have the parameter, we need to configure the node settings.

Steps to Configure the Node

1. Click Properties in the HostedSuite AI Call Insights node.

2. In the Recording Field, select:

   • Flow Parameter: RecordingFile (or the name of your parameter)

3. Enter the HostedSuite API Endpoint:

   • https://v3.hostedsuite.com/api

4. Enter the HostedSuite Customer, Username, and Password.

   • (This information is provided by EvoVoice. Contact support if needed.)

5. Click Save.

Your HostedSuite Insights Flow is now ready to be used in workflows.

Step 3: Connecting the Insights Flow to a Parent Flow

To integrate this subflow into a larger workflow, we need to create an audio recording parameter in the parent flow.

Create an Audio Recording Parameter

1. In the Parent Flow, select the Dial Node.

2. Go to the Parameters tab.

1. Click Add Parameter, then enter:

   • Name: RecordingAudio (or any temporary name to store the recording)

   • Type: AudioFile

2. Click Add Parameter.

This parameter will temporarily hold the recording before sending it to the HostedSuite AI Call Insights node.

Step 4: Configuring the Dial Node for Call Recording

Now, we must enable call recording in the Dial Node.

Steps to Configure Call Recording

1. Open the Dial Node Properties.

1. Update the following settings to Yes:

   • Record Call → ✅ Enable recording

   • Record Transfer Attempts → ✅ Record transfer attempts

   • End Recording After Hangup → ✅ Automatically stop recording when the call ends

2. Click Save.

Step 5: Mapping the Recording to HostedSuite Analytics

Now that both the Parent Flow and AI Call Insights Flow are ready, we need to connect them.

Steps to Map the Recording

1. In the Parent Flow, locate the Recording Ready event.

2. Connect the Recording Ready event to the HostedSuite AI Call Insights subflow.

   • This ensures that the recording is only sent for AI processing when it's ready.

3. Add the HostedSuite AI Call Insights Node to the parent flow.

4. Click Properties and set:

   • Recording File → Flow Parameter: RecordingAudio (the parameter from the Dial Node)

5. Click Save.

Final Setup

With these steps completed, your EvoVoice setup is ready for AI analysis. The system will now process call recordings, apply AI prompts, and generate insights for the Sentinel process.

For further assistance, contact Evo’s support team.