Knowledge Performance Analytics

Review Conversations

• Displays all chatbot conversations across different channels.
• Analysts can review historical chat sessions.
• Provides insights into chatbot compatibility with knowledge sources.

Knowledge Insights & Performance Metrics

• Successful Responses Rate: Measures how often the bot provides an accurate and useful answer to user queries.
• Unrecognized Utterances: Tracks how many users queries the bot failed to respond to due to missing or incorrect knowledge.
• Knowledge Source Utilization: Analyzes which knowledge sources (Documents, SharePoint, External Websites, Knowledge Articles) are being referenced by the chatbot.
• Escalation Trends: Measures how frequently users escalate queries to a live agent due to dissatisfaction with both responses.
• Response Accuracy Analysis: Evaluates how many chatbot responses match the expected answers in the knowledge base.

Identifying & Addressing Knowledge Gaps

• Any failed utterances (user questions that the bot couldn't answer) are recorded under Knowledge Gaps.
• The analyst can review these knowledge gaps and determine if:
  • The knowledge base needs to be updated with new information.
  • Existing knowledge articles require refinements for better accuracy.
  • The bot requires additional training to improve response effectiveness.
• Provides an option to mark an utterance for training, allowing AI models to learn and adapt to user queries more effectively.

Performance Breakdown by Knowledge Category

• The chatbot's knowledge base is divided into categories such as FAQs, Online Forms, Featured Questions, Software Requests, and Hardware Requests.
• Analysts can evaluate which categories are performing well and which require further enhancement.
• This breakdown helps in understanding user preferences and optimizing chatbot responses accordingly.