Conversation Insights Agent Basics
This section describes the basics of using Conversation Insights Agent. It provides an overview of the most important concepts.
Agent
A contact centre agent is a professional who handles customer interactions within a contact center environment. Their primary role is to assist customers via various communication channels, including phone, email, chat, social media, and messaging apps.
Customer
A contact centre customer is an individual or organization that interacts with a contact centre to seek assistance, information, or support related to products, services, or inquiries. These customers reach out through various channels, such as phone calls, emails, live chats, social media, or messaging apps, to communicate with contact centre agents.
Organization Hierarchy
For every Organization in Conversation Insights Agent, there can be multiple hierarchical levels to effectively manage the various teams and departments. Each level of the organization hierarchy is called a Node. You can give access to different Nodes (teams or departments) to use a Fact and Scorecard. When you select a Fact and Scorecard for an organization or its hierarchy (node), the conversation that belongs to that specific node(s) is fetched for processing.
Note
Each node can have up to 7 child nodes.
Conversation
A conversation is an interaction between an agent and a customer in a contact centre. As a data object, it has several properties, including but not limited to:
Date
Session ID
Channel
Direction
Agent
Metadata
Channel
While Conversations can typically be in a variety of communication channels – audio, chat, or email. Conversation Insights Agent supports voice and chat.
Voice: The audio files should be in stereo format and uploaded to Conversation Insights Agent in a .wav format. Currently, mono calls are not supported.
Chat: The chat file should be in plain text with no rich text or HTML tags. Ensure the character set used in the chat file matches that of the ASR-generated transcript for the Voice conversation.
Conversation Processing
When a conversation is processed by Conversation Insights Agent, the conversation recording undergoes a series of audio and text processing services to deliver key metrics and insights.
The call goes through the following audio and text processing services:
Automatic Speech Recognition (ASR): Transcribes the call and identifies speakers (Agent and Customer).
Data Redaction: Redacts PII data of the customer.
Voice Activity Detection: Identifies the duration of customer talk, agent talk, talkover, silence, rate of speech, and call hold count, based on audio of the call.
Tonal sentiment detection: Analyzes both text and tone to identify customer sentiment throughout the call.
Facts: Resolves Facts
In addition, scorecards are evaluated against each call within the respective organization(s) where they are created.
Facts
Facts are atomic building blocks of information that contain a single piece of information that can then be used either individually or together to provide insights into conversations. A Fact uses an LLM to answer questions based on a call transcript. It is a simple and powerful way to derive intelligent insights from conversations.
Scorecards
A contact centre scorecard is a tool to evaluate an agent’s performance based on certain key performance indicators. Scorecards are used for quality assurance, performance evaluation and management, training and development, and customer experience improvement. Scorecards helps agent by providing a clear picture of their performance.
Transcripts
A transcript is a textual record of an audio call. It captures the dialogue between an agent and a customer, and displays each turn with the speaker identified, along with the timestamp of that turn. Transcripts are generated using Uniphore’s in-house Automatic Speech Recognition (ASR) software.
Turn: A turn is a segment of speech in the transcript by a single speaker before the next speaker speaks.
Topics
A topic is a subject spoken by a customer in a conversation.
Manual Quality Management (QM)
In contact centres, Quality Management (QM) is a set of workflows and processes where agents' and teams' performance is continuously measured and monitored so that the business can drive constant improvement.