Dashboard#
The Dashboard allows you to monitor the operational status of your agents in real time and review key performance metrics to maintain stability and continuously improve quality.Verify whether the agent is functioning properly.
Check for any failed calls.
Ensure that token usage during text generation is within a reasonable range.
Detect any irregularities in response quality.
Confirm that feedback is being collected appropriately.
View key data categorized by daily / weekly / monthly intervals.
Analyze performance trends over time to support operational strategy decisions.
Separated Environment DataThe dashboard provides separate analytics for test and live environments, allowing you to clearly distinguish between test results and actual user responses.
For deeper analysis, you can download the full dataset as an Excel file to process externally or include it in internal reports.
Call Count#
Displays the total number of agent calls during the selected period.
Only successful calls are counted toward billing.
Token Usage#
Shows LLM usage records during the aggregation period.
Separates Input Tokens and Output Tokens by model type used.
Agent Quality#
Displays the response quality metrics for the selected period.
Tracks the number of failed responses or those with a confidence score below 0.5.
Worker Activity Log#
Displays the number of cases handled by workers during the aggregation period.For example, if feedback is given on August 16 for a conversation that occurred on August 15, it will be counted under August 15.
Tracks the total number of feedback actions marked as “Good 👍” or “Needs Improvement 👎.”
Modified at 2025-10-20 05:54:40