TABLE OF CONTENTS

Introduction

VoiceOps is a coaching enablement platform that allows your company to quickly assess agent adherence against expected behaviors within calls and provides coaching tools that are used to improve agent performance. 

A key element to implementing the VoiceOps platform is an integration with your call recording solution.  This document provides a brief overview of the VoiceOps platform and outlines our integration capabilities using the VoiceOps API.

How VoiceOps Works

The VoiceOps solution imports call recordings and associated metadata from customer call recording platforms, transcribes the calls into full text, and analyzes that text using Artificial Intelligence (AI) and Machine Learning (ML) to identify key behaviors that reps are or are not adhering to across the calls. The process is supported by a VoiceOps QA team that oversees the model outputs. The output of these analyzed calls is then provided in a web interface for your managers, coaches, and others to assess behavior adherence and to coach agents with the objective of improving that adherence.

VoiceOps API Overview

The VoiceOps API allows customers to quickly and directly integrate their calls into the VoiceOps platform.

The benefits of using the VoiceOps API  are as follows:

  • A quick, simple PUT request records your call's attributes in the VoiceOps system

  • An upload of the associated audio file directly to Amazon's S3 servers connects to your nearest Amazon POP (point of presence), which provides high availability and connection speeds.   

  • These API calls can be performed in a multithreaded fashion, and there's no need for all calls to be blocked on the upload of a single, bulk zip archive.

  • There's no lag time between the raw upload and VoiceOps’ internal transformation of the call. 

Below is a high-level diagram that shows how the API works, allowing you to efficiently record the call attributes in our system, and then upload the audio file directly to Amazon's S3 servers:
Figure 2: VoiceOps Data Integration

Implementing the VoiceOps API

The entire technical documentation can be found at https://api.voiceops.com/ingest/v1/docs

Process

The process to implement the VoiceOps API is as follows:

  1. Review the technical documentation:
     https://api.voiceops.com/ingest/v1/docs

  2. Plan the integration with your internal stakeholders to define:

    1. Which users’ calls will be integrated into VoiceOps.

    2. What metadata, if any, is available and needed for these calls.

Note: VoiceOps can assist with this planning conversation.  Reach out to support@voiceops.com or your Client Success Manager to arrange this meeting.

  1. Determine the Integration_ID value you will use.  The Integration ID is a string that uniquely identifies the rep on this call, which can be viewed and configured at https://app.voiceops.com/v/roster-management. This value must exactly equal the "integration_id" of the expected user on the roster management page.

  2. Contact support@voiceops.com:

    1. Request Customer Admin/API access to your company’s VoiceOps account.

    2. Request the API keys for your account.

    3. Provide the domain(s) of any email address(es) you plan to use for testing the API.

  3. Log into VoiceOps and configure your password:  https://app.voiceops.com/

  4. Access VoiceOps Roster Management: https://app.voiceops.com/v/roster-management

  5. Add or edit the user accounts for which you want to integrate calls.  Documentation on Roster Management can be found here

  6. Test, develop and implement your integration.

FAQ

Q. Do I need to send calls for all users of my call platform?
A. No; only those users whose calls will be transcribed and analyzed by VoiceOps.  Other users need not be integrated. 

Q.  What if there are multiple users in a call file?  For example, the call begins with one user, and then is transferred to another user?
A.  Calls can be attributed to a single user only.  Only the needed call section for the targeted user should be integrated.  

Q.  Can VoiceOps integrate calls that consist of multiple segments/audio files?
A.  Each audio file is considered to be a separate call.  If your call platform splits calls into multiple files, they should be integrated into a single file before sending to VoiceOps.

Q.  I’m having trouble with the API.  Can I reach out to VoiceOps for help?
A.  Yes.  Reach out to support@voiceops.com or your customer success manager and they will find the right resource for your questions.