Evaluation Metrics

Evaluation Metrics is a key component of the Quality AI module that enables supervisors to define, customize, and monitor performance indicators for measuring the quality of agent-customer interactions. The system supports six measurement types, each designed for specific evaluation needs through advanced AI-driven analysis and traditional rule-based methods.

Key Benefits

• AI-Powered Intelligence: GenAI-based adherence reduces dependency on extensive training datasets.

• Comprehensive Coverage: Six measurement types address diverse evaluation scenarios.

• Multilingual Scalability: Enhanced support across different languages and interactions.

• Automated Quality Assurance: Reduces manual review workload through intelligent analysis.

• Real-time Validation: API integration ensures data accuracy and compliance.

• Flexible Configuration: Static and dynamic evaluation options for various use cases.

Access Evaluation Metrics

Access the Evaluation Metrics by navigating to Contact Center AI > Quality AI > Configure > Evaluation Forms > Evaluation Metrics.

Interface Elements

The Interface displays the following elements:

• Name: Shows the evaluation metrics name.

• Metric Type: Indicates the measurement type used.

• Evaluation Forms: Shows all associated evaluation forms for configuration and assignment.

• Ellipsis Icon: Provides edit and delete an options.

• Search: Quick search to view and modify metrics.

• New Evaluation Metrics: Button to configure new metrics.

Create New Evaluation Metrics

Steps to create new evaluation metrics:

1. Select the Evaluation Metrics tab.
   

2. Click the + New Evaluation Metric displayed in the upper-right corner.
   

3. Configure your chosen measurement type metrics.
   

Metrics Configuration Elements

• Metric Naming: Descriptive identifiers for future reference.

• Language Selection: Multilingual support configuration.

• Evaluation Questions: Supervisory reference prompts.

• Adherence Types: Static (universal) vs. Dynamic (trigger-based) detection methods comparison.

Detection Methods Comparison

Feature	GenAI-Based	Deterministic
Mechanism	LLM contextual understanding	Description‑based Semantic similarity matching
Training	Zero-shot prompts	Sample utterance training
Flexibility	High contextual adaptation	Precise pattern recognition
Setup	Description-based configuration	Utterance-based training

Metrics Measurement Types

The following measurement types are:

1. By Question

   Evaluates adherence to specific questions asked or answered during interactions.

   Key Features:

   • Static Adherence: Universal application across all conversations.

   • Dynamic Adherence: Trigger-based conditional evaluation.

   • GenAI Detection: Contextual understanding without training samples.

   • Deterministic Detection: Semantic similarity matching with predefined utterances.

   • Flexible Thresholds: 60% for greetings, 100% for compliance-critical statements.

   Use Cases:

   • Script adherence

   • Greeting compliance

   • Policy verification

   • Response quality assessment

   For the detailed configuration, see By Question.

2. By Speech

   Analyzes speech characteristics and audio quality metrics during voice interactions.

   Key Features:

   • Cross Talk: Monitors simultaneous speaking instances with customizable thresholds and duration limits.

   • Dead Air: Tracks unproductive silence periods with configurable thresholds (30-300 seconds).

   • Speaking Rate: Monitors speech pace trends to flag coaching opportunities, measuring words per minute (WPM).

   Use Cases:

   • Voice interaction quality

   • Conversation flow analysis

   • Speaking pace optimization

   For the detailed configuration, see By Speech.

3. By Value

   Verifies agent-shared customer-specific information against trusted data sources.

   Key Features:

   • API Integration: Real-time verification with CRM and external systems.

   • Business Rules Engine: Five rule types, including first or last value, negotiated value, and strict matching.

   • Compliance Tracking: Automated deviation detection for regulatory requirements.

   • Audit Trails: Detailed documentation for supervisory review.

   Use Cases:

   • Pricing accuracy

   • Interest rate verification

   • Account balance confirmation

   • compliance validation

   For the detailed configuration, see By Value.

4. By Dialog Task

   Evaluates completion and quality of specific dialog tasks and workflows.

   Key Features:

   • Dialog Agent Selection: Choose from available dialog agents.

   • Evaluation Scope: Entire conversation or time-bound assessment.

   • Time Parameters: Configurable seconds (voice) or messages (chat).

   Use Cases:

   • Workflow adherence

   • Task completion verification

   • Dialog flow optimization

   For the detailed configuration, see By Dialog Task.

5. By Playbook Adherence

   Assesses compliance with predefined agent playbooks and procedures.

   Key Features:

   • Entire Playbook: Comprehensive adherence across all playbook elements.

   • Specific Steps: Targeted evaluation of particular stages and steps.

   • Percentage Thresholds: Configurable minimum adherence requirements.

   Use Cases:

   • Process compliance

   • Procedure adherence

   • Standardization enforcement

   For the detailed configuration, see By Playbook Adherence.

6. By AI Agent

   Enables sophisticated evaluations using AI agents capable of multi-step reasoning and autonomous decision-making.

   Key Features:

   • Complex Analysis: Multi-step reasoning connecting conversation elements.

   • Domain Expertise: Specialized knowledge requirements (compliance, technical support).

   • Contextual Understanding: Nuanced evaluation requiring full conversation context.

   • Advanced Decision-Making: Sophisticated judgment calls beyond pattern matching.

   Used Cases:

   • Complex compliance assessments

   • Technical troubleshooting evaluation

   • Sophisticated quality analysis

   For the detailed configuration, see By AI Agent.

Manage Evaluation Metrics

The process of managing evaluation metrics includes the following sections:

Metric Lifecycle Management

• Create: Step-by-step configuration wizard

• Edit: Real-time updates with validation

• Delete: Dependency resolution and cleanup

• Language Management: Multilingual configuration and updates

Dependency Management

• Form Associations: Evaluation form linkage requirements

• Attribute Assignments: Metric-to-attribute mapping

• Language Dependencies: Active language protection

• Deletion Prerequisites: Dependency resolution before removal

Edit or Delete Evaluation Metrics Type

Steps to edit or delete existing evaluation metrics types:

1. Right-click on any of the existing Evaluation Metrics Type.
   

2. Click Edit to to update measurement type fields or Delete to remove.
   

3. Click Update to save changes.

Send Feedback