Evaluation Scores
Overview
The Evaluation Scores section provides comprehensive quality assessment and performance scoring for your AI agents. This guide covers everything from automated quality scoring to user feedback integration, helping you maintain and improve agent performance through data-driven insights.
1. Scores Dashboard
Evaluation Scores Interface - Monitor quality metrics and performance scores
The scores dashboard displays:
- Overall Quality Score: Aggregate quality across all agents
- Performance Metrics: Speed, accuracy, and reliability scores
- User Satisfaction: Direct feedback and ratings from users
- Trend Analysis: Score changes over time
- Comparative Rankings: How agents perform relative to each other
📸 Screenshots
To add screenshots:
- Navigate to the Evaluation section in your app
- Take screenshots of the relevant interfaces
- Save them in
docs/public/images/screenshots/evaluation/ - Run
python3 scripts/integrate_screenshots.pyto automatically integrate them
Recommended screenshots:
scores-view.png- Performance scores view
Tip: Use PNG format with 1920x1080 resolution for best results.
Overview
The Performance Scores sub-section is a part of the larger Evaluation section in Signpost AI. It is designed to provide a comprehensive analysis of agent performance through various metrics and scoring systems. This feature is crucial for managers and supervisors who need to monitor and evaluate the performance of their team members.
Users would typically use this feature to:
- Monitor individual agent performance
- Compare performance across different agents
- Identify trends and patterns in performance data
- Make data-driven decisions on agent performance improvement
Key Features
Score Calculation
This feature calculates the performance score of each agent based on various metrics such as call handling time, customer satisfaction, and resolution rate. To access this feature, navigate to the 'evaluation/scores.tsx' file.
Performance Metrics
Performance metrics provide a detailed breakdown of each agent's performance. These metrics can be accessed in the 'evaluation/score.tsx' file.
Comparison Tools
Comparison tools allow users to compare the performance of different agents. This feature can be found in the 'evaluation/scores.tsx' file.
Trending Analysis
Trending analysis helps users identify patterns and trends in agent performance over time. This feature is available in the 'evaluation/score.tsx' file.
User Interface
The Performance Scores sub-section is designed with a clean and intuitive interface. The main screen displays a list of agents along with their performance scores. Users can navigate to individual agent profiles to view detailed performance metrics.
Interactive elements include dropdown menus for selecting different metrics and comparison tools, and a date range picker for trending analysis.
How to Use
To use the Performance Scores sub-section:
- Navigate to the Evaluation section in Signpost AI.
- Click on the Performance Scores sub-section.
- View the overall scores of all agents.
- Select an agent to view detailed performance metrics.
- Use the comparison tools to compare performance across different agents.
- Use the trending analysis feature to identify performance trends over time.
Configuration & Settings
Users can customize the Performance Scores sub-section by selecting the metrics they want to display. Advanced configuration options include setting custom scoring algorithms and defining the weightage of different metrics in the overall score calculation.
Integration
The Performance Scores sub-section integrates with the Call Handling and Customer Satisfaction sections of Signpost AI. It uses data from these sections to calculate performance scores and metrics.
Troubleshooting
Common issues include incorrect score calculations and missing performance data. If you encounter these issues, ensure that all data inputs are correct and that the system is correctly configured. If the problem persists, contact the Signpost AI support team.