Conversation History

The Conversation History page in Symphona provides a clear overview of all interactions between users and the AI agents. This page enables you to track, review, and analyze individual conversations in detail, making it easy to evaluate AI performance, user satisfaction, and identify opportunities for improving customer experience.

To reach this page, in the sidebar select Converse > Conversation History.

The Conversation History page displays a list of individual conversation logs with high-level information clearly presented for quick reference. Each conversation log entry includes:

• Name: Identifies the user involved in the conversation.
• Experience: Indicates user sentiment (e.g., Positive, Neutral, Negative) based on AI insight into the conversation.
• Related Agents: Displays the AI agent involved in the conversation.
• Closed By: Shows who ended the conversation (e.g., User or Agent).
• Conversation Duration: Total duration from start to finish.
• Start Time / End Time: Timestamps clearly indicating when the conversation began and ended.
• Channel Type: The type of platform used.
• Channel Name: Specific identification of the conversation source.
• Location: Geographic location of the user.
• Email / Phone Number: Contact details provided by the user.

Detailed Sidebar View

To examine detailed information for any listed conversation, select a conversation entry from the history list. A detailed sidebar will open, displaying extensive information on the conversation. The sidebar provides an in-depth breakdown of individual conversations, including interaction transcripts, conversation details, user information, and feedback:

Conversation Transcript

A complete, chronological view of messages exchanged during the conversation, clearly indicating:

• Time Stamps: Precise times each message was sent and received.
• System Actions: Automated actions such as "Query Knowledge Base" are clearly highlighted to differentiate them from regular messages.
• Status Updates: Important notifications (e.g., "Conversation marked as closed") appear in distinct colors or sections to clearly indicate the conversation lifecycle.

This transcript helps you accurately reconstruct and evaluate the context, quality, and flow of user-agent interactions.

Conversation Details Section

The sidebar provides comprehensive metadata regarding the selected conversation:

• Closed By: Clearly identifies who closed the conversation (e.g., End User, Agent, System).
• Conversation ID: A unique identifier to track or reference the conversation later.
• Start Time & End Time: Precise timestamps marking when the interaction began and ended.
• Channel Type: Indicates the communication method used (e.g., Web Chat).
• Channel Name: Clearly identifies the specific widget or communication channel.
• Location: Displays the geographical location of the end-user.
• Source Agent: The AI agent responsible for managing or interacting in the conversation.

This information helps clearly document interactions and supports audits or user support follow-ups.

User Information Section

Clearly presents the user's contact details and identifying information, including:

• Phone Number: User-provided phone contact.
• Email: Email address provided by the user.
• User ID: Identifier used to track user interactions within the system. User ID consist of one of the following:
• Symphona ID if they are testing within the app
• ID provided through the start conversation REST API
• ID provided through the chat widget setUserContext() functions
• Nothing

This data allows for effective user follow-up and personalized customer support.

User Experience & Feedback Section

Provides insights into the quality of the user experience:

• Conversation Experience (AI Insight):
• Positive / Neutral / Negative: An AI-generated sentiment rating summarizing overall user experience quality based on message analysis and interactions. Quickly identifies user satisfaction levels.
• How was this calculated?: A clear explanation of how AI determined the experience rating, including interaction content, sentiment, or methodology.

The User Experience & Feedback section helps identify strengths and areas for improvement, driving continuous enhancement of your AI interactions.

How to Effectively Use the Conversation History Page:

• Monitor Customer Satisfaction: Regularly review "Experience" ratings to quickly identify problematic conversations and investigate further.
• Audit AI Performance: Check conversation transcripts to see if the agent provided accurate responses or if improvements to training data are required.
• User Engagement: Use detailed contact information provided to proactively reach out to users if necessary.

This Conversation History page gives your team powerful tools for managing, reviewing, and optimizing user-agent interactions, significantly enhancing user experience management.

FOA Dashboard

The Front Office Automation Dashboard provides real-time insights into customer interactions handled by your automated front office channels. The displayed metrics and charts will update based on the selected date range and process, allowing you to focus on the data that matters most.

To access this dashboard in the Symphona app, navigate on the left side menu to FOA Dashboard. By default, the Conversations tab is active, and you can click on Service Tickets to switch tabs.

Conversations

The Conversations tab provides an overview of how your agents are running within the selected timeframe. It shows you how many conversations have occurred, their transferred rate, and average conversation lengths. This tab helps you answer questions like: How often are there conversations? Are they being transferred? What is the average length of a text conversation?

Conversations Overview

In the Conversations Overview section, you will see summary cards with key metrics about conversations. These metrics give a quick health check of your agents:

• Total Conversations: The total number of conversations in the chosen timeframe. Below are specifics on whether they were voice or text conversations.
• Total Conversations Transferred: The total number of conversations transferred in the chosen timeframe. Below are specifics on whether they were voice or text transferred conversations.
• Average Voice Conversation Length: How long the average voice conversation lasted.
• Average Text Conversation Length: How long the average text conversation lasted.

Conversations

Below the overview, the dashboard visualizes conversations over time. This section includes a chart that show when and how many conversations are occurring.

Key features of the Process Execution Performance chart include:

• Time Distribution Chart: A chart that displays the number of conversations over time within the selected period. User’s can choose between a bar or line chart by pressing the settings button, and whether the graph should display voice or text conversations. The x-axis represents time intervals (for example, hours of the day if viewing Today, or days/weeks if viewing a longer range), and the y-axis shows the number of conversations.
• Frequency Selection: You can adjust the granularity of the chart. For short timeframes like a single day, the chart may default to an hourly view (each bar = one hour). For longer timeframes, it might show daily totals. In some cases, the dashboard lets you switch the frequency (e.g., view by hour, day, or month) to get the appropriate level of detail. Users can also view more detailed frequency selections by moving the orange bar at the bottom.

Average Conversation Length

The next dashboard visualizes average conversation lengths over time. This section includes a chart that show when and how long the average conversations are.

Key features of the Average Conversation Length chart include:

• Time Distribution Chart: A chart that displays the length of conversations, in minutes, over time within the selected period. User’s can choose between a bar or line chart by pressing the settings button, and whether the graph should display voice or text conversations. There also options to view the moving average as well as seeing individual channels through separate lines.
• Frequency Selection: You can adjust the granularity of the chart. For short timeframes like a single day, the chart may default to an hourly view (each bar = one hour). For longer timeframes, it might show daily totals. In some cases, the dashboard lets you switch the frequency (e.g., view by hour, day, or month) to get the appropriate level of detail. Users can also view more detailed frequency selections by moving the orange bar at the bottom.

Transferred Conversations

At the bottom of the tab, the dashboard visualizes the number of transferred conversations over time. This section includes a chart that show when and how conversations are transferred.

Key features of the Average Conversation Length chart include:

• Time Distribution Chart: A chart that displays the length of conversations, in minutes, over time within the selected period. User’s can choose between a bar or line chart by pressing the settings button, whether the graph should display numbers or percentages, and whether the graph should display voice or text conversations. There also options to view the moving average as well as seeing individual channels through separate lines.
• Frequency Selection: You can adjust the granularity of the chart. For short timeframes like a single day, the chart may default to an hourly view (each bar = one hour). For longer timeframes, it might show daily totals. In some cases, the dashboard lets you switch the frequency (e.g., view by hour, day, or month) to get the appropriate level of detail. Users can also view more detailed frequency selections by moving the orange bar at the bottom.

• Conversation History
• Detailed Sidebar View
• Conversation Transcript
• Conversation Details Section
• User Information Section
• User Experience & Feedback Section
• How to Effectively Use the Conversation History Page:
• FOA Dashboard
• Conversations
• Conversations Overview
• Conversations
• Average Conversation Length
• Transferred Conversations