Exports

The export functionality allows users to generate comprehensive reports based on the data from conversations, customers, and agents. The system supports two types of exports: a Generic Export for aggregated data, and a Single Conversation export, a focused, record-by-record analysis of an individual conversation.

• Generic export 
• Single Conversation Export

Generic export 

Exports can be generated only for filtered data based on user-defined filters. Unlike for other datastreams, scheduled export is not available for conversations. 

The export includes the following sheets:

Conversation 

The purpose of this sheet is to offer a summary of each conversation, along with its associated metadata and analysis results.

• This sheet contains conversations and aggregates data by conversation.
• Includes conversation-level metadata (e.g., conversation length, customer ID, agent ID)
• Results of text analysis at the conversation level, including:
  • Aggregate sentiment scores Overall sentiment, Starting sentiment, Ending sentiment
  • Recognized Custom Labels (CLs) for the entire conversation
  • Reason of contact 

Messages 

The purpose of this sheet is to allow users to view each message in its full detail with associated metadata and analysis results.

• This sheet contains all individual messages in the dataset, one message per row.
• Metadata for each message (e.g., message timestamp, customer ID, agent ID, etc.)
• Results of text analysis, including:
  • Sentiment score
  • Recognized Custom Labels

Customer 

The purpose of this sheet is to provide a summary of each customer’s interactions, including key insights based on the customer’s messages and text analysis results.

• This sheet aggregates data by customer ID.
  • Customer-level metadata (e.g., total number of conversations, average conversation length)
  • Results of text analysis aggregated by customer, including:
    • Average sentiment score
    • SUM sentiment score
    • All recognized Custom Labels (CLs)
    • Reason of contacts in conversations

Agent 

The purpose of this sheet is to provide a summary of each agent’s performance, including insights based on the messages they’ve handled.

• This sheet aggregates data by agent ID.
  • Agent-level metadata (e.g., total number of conversations handled, average conversation length)
  • Results of text analysis aggregated by agent, including:
    • Average sentiment score
    • All recognized Custom Labels (CLs) associated with the agent’s conversations
    • All recognized reason of contacts associated with the agent’s conversations

Single Conversation Export

The Single Conversation Export is designed for users who require a detailed export of one specific conversation. Available directly from the extended view of a conversation, this export includes every message within the conversation, as well as all related metadata and text analysis results.

The Single Conversation Export includes the following sheets:

Conversation 

• Contains all messages from the specific conversation, listed message by message.
• Included Data:
  • Message-level metadata (e.g., message timestamp, customer ID,customer name, agent ID, agent name)

Metadata 

• Contains the aggregated metadata for the entire conversation.
• Included Data:
  • Conversation-level metadata (e.g., conversation ID, customer ID, agent ID, total duration)
  • Starting sentiment, closing sentiment, reason of contact 
  • Any other metadata provided alongside the conversation

Customer Text Analytics 

• Structure: Contains all messages from the customer, listed row by row.
• Included Data:
  • Message-level metadata (e.g., message timestamp)
  • Results of text analysis for each customer message, including sentiment score, positive/negative phrases, and recognized Custom Labels.

Agent Text Analytics 

• Structure: Contains all messages from the agent, listed row by row.
• Included Data:
  • Message-level metadata (e.g., agent ID, agent name)
  • Results of text analysis for each agent message, including sentiment score, positive/negative phrases, and recognized Custom Labels (CLs).