Sairis

Appearance

AI Assistant Dashboard Guide ​

Introduction ​

The AI Assistant Dashboard provides administrators and managers with detailed analytics on AI usage across your organization. This specialized dashboard focuses on monitoring AI performance metrics, tracking conversation statistics, and analyzing token consumption to help optimize your AI deployment and manage costs effectively.

Accessing the AI Assistant Dashboard ​

The AI Assistant Dashboard is accessible to users with the following roles:

  • Admin
  • Content
  • Manager

To access the dashboard, navigate to the Insights section from the main navigation menu and select "AI Assistant Dashboard."

Dashboard Interface Overview ​

The AI Assistant Dashboard is designed to focus specifically on AI conversation analytics and resource usage across your tenant. The dashboard's primary component is the Token Consumption Analysis chart.

Token Consumption Analysis ​

The main chart in the dashboard displays tenant-wide token usage statistics:

Understanding Token Usage ​

The chart provides a comprehensive view of:

  • Input Tokens: Number of tokens used in user prompts and questions
  • Output Tokens: Number of tokens generated in AI responses
  • Combined Usage: Total token consumption across your organization

Time Period Analysis ​

The chart allows you to analyze token usage across different time periods:

  • Daily View: Shows usage patterns by day
  • Weekly View: Aggregates usage by week
  • Monthly View: Shows monthly consumption trends

Visualization Features ​

The token consumption chart includes several helpful visualization features:

  • Color-Coded Lines: Different colors for input, output, and total tokens
  • Hover Details: Precise numbers when hovering over data points
  • Interactive Legend: Toggle specific metrics on/off
  • Date Range Selection: Analyze specific time periods

Using the Dashboard for Resource Management ​

Cost Monitoring ​

Token usage directly correlates to AI service costs. The dashboard helps you:

  • Track overall consumption trends
  • Identify unusually high usage periods
  • Plan for expected usage increases
  • Set appropriate credit allocations for users

Usage Pattern Analysis ​

By examining the ratio of input to output tokens, you can:

  • Understand how users are interacting with the AI
  • Identify potential inefficiencies in prompt structure
  • Detect patterns that might indicate suboptimal usage

Optimizing AI Deployments ​

The dashboard helps you make data-driven decisions about:

  • Which AI models provide the best balance of quality and cost
  • How to structure prompts more efficiently
  • When to recommend Blueprint usage over ad-hoc conversations

Interpreting Dashboard Data ​

Normal Usage Patterns ​

Typical healthy usage patterns show:

  • Consistent day-to-day or week-to-week usage
  • A balanced ratio between input and output tokens
  • Gradual increases as more users adopt the system

Notable Anomalies ​

Watch for these patterns that may require investigation:

  • Sudden spikes in token usage
  • Unusually high output token counts relative to input
  • Dramatic drops in usage (may indicate system issues)

Sharing Dashboard Insights ​

To share insights from the AI Assistant Dashboard:

  1. Use browser screenshot functionality to capture the current view
  2. Export specific metrics for inclusion in reports
  3. Schedule regular reviews with stakeholders to discuss usage trends and cost implications

Additional Dashboard Functionality ​

Depending on your implementation, the AI Assistant Dashboard may also include:

  • User-Level Metrics: Drill down to see which users or departments use the most tokens
  • Blueprint Comparisons: Compare token efficiency between different Blueprint configurations
  • Content Source Analysis: Identify which knowledge bases generate the most token usage
  • Cost Projections: Estimate future costs based on current usage trends

Troubleshooting ​

Issue: Chart shows no data

  • Verify AI conversations have occurred in the selected time period
  • Check if your role has permission to view organization-wide metrics
  • Try extending the date range to capture more historical data

Issue: Unexpected peaks in token usage

  • Investigate if new users were onboarded during that period
  • Check if any large-scale AI projects occurred during that time
  • Verify no unauthorized or unusual access patterns occurred

Issue: Cannot access certain dashboard features

  • Some detailed analytics may require specific administrative permissions
  • Check with your system administrator about role-based access controls
  • Ensure your account has the necessary analytics permissions enabled

For additional assistance with the AI Assistant Dashboard, please contact your system administrator.