Transcript Filtering
Transcript filtering for a chatbot with parameters like "recorded in," "recorded after," "recorded before," "longer than," and "shorter than" is a valuable feature that enables users to selectively retrieve and display chatbot conversations or sessions based on specific criteria. The filtered data lists these conversations along with various details, including Thread, Duration, User Messages, Virtual Agent Messages, Human Agent Messages, Escalated status, and Handled status. Users can click on any conversation or session to access and view the full conversation details.
Key Components and Significance
Filtering Predicates
Recorded In
Users can filter conversations that occurred within a specified timeframe, such as a specific date, month, or year.
Use case: Users can narrow down conversations to a particular time period to analyze trends or incidents within that timeframe.
Recorded After
Conversations recorded after a specific date and time can be filtered using this parameter.
Use case: Useful for isolating recent conversations to assess recent user interactions or changes in chatbot performance.
Recorded Before
Users can filter conversations recorded before a specified date and time.
Use case: Helps in reviewing historical data or comparing current interactions with past ones.
Longer Than
Conversations with a duration longer than a defined threshold, often specified in minutes or seconds, can be filtered using this parameter.
Use case: Useful for identifying lengthy conversations that may require special attention or analysis.
Shorter Than
Conversations shorter than a specified duration can be filtered using this parameter.
Use case: Allows users to identify brief interactions that may require a different analysis or follow-up.
Filtered Data Display Details Provided
After applying the desired filters, the chatbot analytics platform lists the filtered conversations or sessions.
Thread
A reference or identifier for the conversation or session, which allows users to distinguish between different interactions.
Duration
The duration of the conversation or session, indicating how long the interaction lasted.
User Messages
The count of messages or inputs provided by the user during the conversation or session.
Virtual Agent Messages
The count of messages or responses generated by the chatbot or virtual agent.
Human Agent Messages
The count of messages sent by a human agent, if applicable.
Escalated
Indicates whether the conversation or session was escalated to a human agent or a higher level of support.
Handled
Indicates whether the conversation or session was successfully handled or resolved.
Clickable Threads
Users can click on any conversation or session entry in the filtered list to access and view the complete details of that specific conversation or session.
Significance and Use Cases
Performance Analysis
Users can analyze chatbot performance during specific time periods using the "Recorded In," "Recorded After," and "Recorded Before" filters.
Quality Control
Monitoring conversations longer or shorter than typical durations can help identify quality issues, user engagement patterns, and opportunities for improvement.
Incident Investigation
Filtering conversations by date and time is crucial for investigating specific incidents, tracking user interactions, and identifying trends.
Trend Analysis
Users can use filtering to track how conversation lengths, escalation rates, and handling success rates change over time.
User Support
Support teams can prioritize handling conversations that were escalated or remain unresolved.
Summary
In summary, transcript filtering for a chatbot with parameters like "recorded in," "recorded after," "recorded before," "longer than," and "shorter than" provides users with a comprehensive tool for exploring, analyzing, and managing chatbot interactions. The inclusion of details such as duration, message counts, escalation status, and handling status enhances the depth of insights and enables effective performance monitoring and improvement of chatbot interactions.