Mastering the Question Base Dashboard: Your Ultimate Guide to AI Agent Performance

The Question Base dashboard serves as your central command center for monitoring and optimizing your AI Answer Agent's performance. Located at questionbase.ai, this powerful interface provides a comprehensive overview of how your AI is handling team inquiries and highlights areas that need your attention.

When you navigate to the Dashboard view from the left side of the screen, you'll immediately see a breakdown of your agent's performance metrics. This at-a-glance view helps you understand how effectively your AI is answering questions across your organization's Slack channels.

The dashboard is designed to guide your continuous improvement efforts by categorizing responses into helpful answers, unhelpful answers, and instances where no answer was found. This organization makes it simple to prioritize your optimization efforts and ensure your AI continues to deliver value to your team.

Let's explore each section of the dashboard to understand how you can leverage these insights to enhance your AI agent's performance and expand your team's knowledge base.

Navigating the Helpful Answers Section to Improve Team Knowledge

The Helpful Answers section of your Question Base dashboard showcases all the AI-generated responses that have been upvoted by your team members. This collection represents the successful knowledge transfers happening across your organization and provides valuable insights into what's working well.

When you click on a specific question in this section, such as "Where do we store the operational documents regarding anything HR and new employees?", you can review the complete answer provided by your AI agent. This gives you visibility into which knowledge sources are being effectively utilized and which topics your team frequently needs assistance with.

For each helpful answer, you'll see an "Open in Slack" option that allows you to view the original context of the question and response. This feature is particularly useful for understanding the full conversation thread and seeing how team members engaged with the answer.

By regularly reviewing the Helpful Answers section, you can:

• Identify commonly asked questions that might benefit from additional documentation

• Recognize patterns in the types of information your team frequently seeks

• Validate that your knowledge base contains accurate and up-to-date information

• Celebrate wins where your AI agent is successfully supporting your team

This section serves as a positive reinforcement mechanism, showing you where your AI agent is excelling and helping you build upon those successes.

Managing Unhelpful Answers to Enhance AI Response Quality

The Unhelpful Answers section of your dashboard is perhaps one of the most valuable for ongoing improvement. This area displays all AI-generated responses that team members have dismissed or voted down, indicating they didn't receive the information they needed.

Each entry in this section represents an opportunity to enhance your AI agent's knowledge base and response quality. For example, when clicking on a question like "How can I get a business card?", you can review exactly what the AI responded with and which knowledge sources it referenced.

This detailed view allows you to pinpoint where things went wrong:

• Was the information outdated?

• Did the AI reference the wrong source?

• Was the answer incomplete or lacking necessary detail?

• Did the AI misunderstand the question intent?

Armed with these insights, you can take targeted actions to improve future responses:

1. Update the referenced knowledge sources with more accurate information

2. Add new documentation to cover gaps in your knowledge base

3. Refine existing content to be more clear and comprehensive

4. Adjust how information is organized to improve retrieval accuracy

By systematically addressing the issues identified in the Unhelpful Answers section, you'll continuously enhance your AI agent's ability to provide valuable responses, ultimately reducing the number of entries in this category over time.

Addressing Knowledge Gaps with the No Answer Found Category

The No Answer Found category highlights instances where your AI agent was unable to provide any response to a team member's question. This section serves as a crucial to-do list for expanding your knowledge base and addressing blind spots in your documentation.

When you explore questions in this category, such as "How can I request a demo account for an Enterprise customer?", you're identifying completely new topics that need to be documented and integrated into your knowledge base.

For each unanswered question, you have two primary options:

1. Integrate new knowledge sources: Create and upload documentation that addresses the topic to ensure future questions on this subject receive accurate answers.

2. Provide an answer directly in Slack: Click "Open in Slack" to respond to the question in the original channel. The AI agent will capture your response and automatically save it to provide similar answers in the future.

This second approach is particularly powerful because it creates a virtuous cycle - as team members ask questions and receive answers (either from the AI or from colleagues), the system continuously learns and expands its knowledge base.

The No Answer Found category effectively functions as an automated knowledge gap analysis, highlighting exactly where your documentation is missing information that your team needs. By regularly reviewing and addressing these gaps, you ensure your AI agent becomes increasingly comprehensive in its ability to support your organization.

Maximizing Agent Performance with Channel-Specific Analytics

The Question Base dashboard offers powerful filtering capabilities that allow you to analyze AI agent performance across different Slack channels and time periods. This granular view helps you identify patterns and prioritize improvements where they'll have the greatest impact.

When you select the channel filter, you can focus on specific areas of your organization. For example, by selecting "#ask-payment," you can see how effectively your AI agent is handling payment-related inquiries specifically. This targeted analysis reveals:

• Which channels have the highest volume of questions

• Where your AI agent is performing particularly well or struggling

• Topic-specific knowledge gaps that need addressing

• Teams that might benefit from additional documentation or training

The time period filters allow you to track improvements over time and measure the impact of your knowledge base enhancements. After updating documentation or adding new sources, you can monitor subsequent weeks to see if the number of unhelpful or missing answers decreases in the relevant channels.

For maximum effectiveness, consider establishing a regular review cadence:

1. Weekly reviews of high-priority channels

2. Monthly analysis of overall trends across all channels

3. Quarterly deep dives into persistent knowledge gaps

By leveraging these channel-specific analytics, you can take a strategic approach to improving your AI agent's performance, focusing your efforts where they'll deliver the greatest benefits to your team.

The Question Base dashboard transforms AI agent management from a reactive process to a proactive one, giving you the insights needed to continuously enhance your team's knowledge sharing and support experience. By regularly engaging with each section of the dashboard, you'll ensure your AI agent becomes an increasingly valuable resource for your entire organization.`