Skip to main content

Dashboard Guide

The Teckel AI dashboard is your central hub for monitoring and improving the quality of your AI systems. This guide will walk you through the main sections of the dashboard and how to use them.

Dashboard​

The Dashboard provides a comprehensive overview of the queries being sent to your AI system and their corresponding audit scores. This is a great way to monitor the overall health of your document base to understand how the chatbot is generally performing.

Trace History​

The Trace History page provides a comprehensive log of all the traces that have been sent to Teckel AI. For each trace, you can see:

  • The user's query and the AI's response
  • Quality Metrics (expandable for technical teams):
    • Response accuracy (accuracy score)
    • Documentation relevance (precision)
    • Answer completeness (completeness)
    • Information currency (document freshness)
  • Impact Analysis: Specific business risks identified (e.g., "Outdated pricing info", "Missing compliance disclaimer")
  • The source documents that were used to generate the response

Documents Analytics​

The Documents section shows you the health of your knowledge base, use this to determine which documents are pulling their weight or might need a touch up.

Technical Metrics​

  • Document Health Scores: Aggregated performance metrics for each document
  • Precision Tracking: How often chunks actually support factual claims
  • Freshness Monitoring: Track document age and staleness
  • Usage Analytics: Frequency and performance of document references

Topics​

The Topics section transforms your AI interactions into actionable intelligence about your documentation. This powerful feature automatically discovers patterns in user queries and identifies precisely where your knowledge base needs attention.

Topic Intelligence​

Our advanced topic discovery system automatically groups similar queries to reveal:

  • Query Patterns: Understand what users are really asking about, with topics automatically named and organized
  • Topic Performance: See average quality scores for each topic to identify areas where your AI consistently struggles
  • Query Volume: Track which topics generate the most questions, helping you prioritize documentation efforts
  • Topic Relationships: Visualize connections between related topics to understand the full context of user needs

How Topic Clusters Form​

Our ML-powered system continuously analyzes incoming queries to:

  • Group semantically similar questions, even with different wording
  • Identify emerging topics as user needs evolve

For example, these queries automatically cluster into "Account Recovery":

  • "I forgot my password, how do I reset it?"
  • "Can't access my account, need help"
  • "Reset password link expired, what now?"
  • "Account locked after failed login attempts"

Knowledge Gap Detection​

Beyond simple clustering, we identify critical gaps in your documentation for each Topic:

  • Documentation Gaps: Topics where AI responses consistently have low accuracy scores, indicating missing or inadequate documentation
  • Precision Issues: Areas where your AI retrieves irrelevant information, suggesting documentation organization problems
  • High-Impact Opportunities: Gaps affecting the most queries, helping you maximize the impact of documentation improvements

Document Performance by Topic​

Understand exactly which documents support which topics:

  • Topic Coverage: See which documents are referenced and support each topic cluster
  • Gap Identification: Identify topics that lack proper documentation coverage
  • Improvement Targeting: Focus updates on documents that affect high-volume, low-performance topics

Document Feedback Consolidation​

When a document repeatedly causes issues, we aggregate the feedback:

Impacts These Topics​

See exactly which topic clusters are affected when a document has problems:

  • Only shows topics where the document received negative feedback
  • Helps understand the blast radius of documentation issues
  • Prioritizes fixes based on affected query volume

For example, if your "API Rate Limits" document has issues, it might impact:

  • "API Performance Optimization" (234 queries affected)
  • "Error Handling" (156 queries affected)
  • "Scaling Best Practices" (89 queries affected)

Actionable Recommendations​

Every Topic and Document comes with specific actions for your subject matter experts:

  • Priority Rankings: Know exactly which documentation gaps to address first
  • Sample Queries: See real user questions that struggled due to documentation issues
  • Performance Trends: Track whether topics are improving or declining over time
  • Impact Estimates: Understand how many queries would benefit from each improvement

This comprehensive analysis empowers your team to move from reactive troubleshooting to proactive documentation quality management, ensuring your AI always has accurate, up-to-date information to draw from.

Administrative Controls​

For users with admin privileges, the dashboard provides a set of administrative controls for managing your organization's settings:

  • API Keys: Generate and manage API keys for your different environments.
  • User Management: Invite new users to your organization, assign roles, and manage team memberships (coming soon).
  • Billing: Review your usage-based billing and company plan.
  • Settings: Configure notifications, set quality score thresholds, and customize other preferences for your organization.