> ## Documentation Index
> Fetch the complete documentation index at: https://koreai-content-gov.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Agent Scorecards

Agent Scorecards evaluate individual agents’ performance across voice and chat interactions using configurable, attribute-based metrics linked to key areas such as sales proficiency, compliance, and product expertise. This supports multilingual evaluations, assigns scorecards by contact direction (inbound or outbound), applies channel-specific rules, and supports negative weight penalties, fatal criteria scoring, and versioned configurations.

You can set a minimum duration threshold to exclude short or incomplete contacts from AutoQA scoring and quality metrics. Supervisors can track automated and manual audit outcomes across dashboards, reports, and APIs, ensuring accurate and consistent performance evaluation across all contact flows.

## Key Capabilities

* **Individual Assessment**: Each agent receives evaluations based on applicable scorecards during call transfers.
* **Direction-aware Assignment**: Assign separate scorecards for Inbound and Outbound contacts to evaluate each direction using relevant criteria.
* **Multiple Evaluations**: Apply multiple scorecards to each interaction to assess agents.
* **Multilingual Support**: Attributes and scorecards support multilingual evaluations for diverse interactions.
* **Negative Weight Scoring**: Apply negative weights to key metric results to highlight their significance and impact on agent scores.
* **Fatal Criteria Parameters**: Mark non-compliance metrics as fatal, resulting in a zero score for the interaction.
* **Not Applicable Scoring Treatment**: Choose whether to exclude metrics marked Not Applicable from scorecard scoring or award them full weightage, independently of the evaluation form's setting.
* **Minimum Duration Threshold**: Exclude contacts that fall below a configured duration from AutoQA scoring.
* **Enterprise Quality Monitoring**: Supports advanced enterprise scoring, reporting, and audit workflows.

***

## Access Agent Scorecards

Navigate to **Quality AI** > **Configure** > **Agent Score Cards**.

<img src="https://mintcdn.com/koreai-content-gov/zzMEkrFkk9Ro74Zq/ai-for-service/quality-ai/configure/agent-scorecards/images/agent-scorecard-landing-page.png?fit=max&auto=format&n=zzMEkrFkk9Ro74Zq&q=85&s=3ae4585919f4315cfde969c8a41c7365" alt="Agent Score Card" width="1909" height="942" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/agent-scorecard-landing-page.png" />

<Note>To view Agent Scorecards and Agent Attributes, enable the **Agent Scorecards** toggle in **Quality AI > Settings > Quality AI Configuration Settings**.</Note>

## Permission Types

Supervisors can have one of three permission levels for agent scorecards:

| Permission      | Description                                         |
| --------------- | --------------------------------------------------- |
| **Full Access** | Create, edit, and manage scorecards and attributes. |
| **No Access**   | Can't create or edit scorecards and attributes.     |
| **View Only**   | Read existing scorecards and attributes, no edits.  |

***

## Create a New Agent Scorecard

Creating an agent scorecard involves the following three sections:

1. Select **+ New Agent Score Card**.

   <img src="https://mintcdn.com/koreai-content-gov/uQy_3EJi2vwOWgyh/ai-for-service/quality-ai/configure/agent-scorecards/images/agent-scorecard-default-page.png?fit=max&auto=format&n=uQy_3EJi2vwOWgyh&q=85&s=3ba30c6d04ee731c7544c36dab958e7b" alt="New Agent Scorecard" width="1100" height="342" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/agent-scorecard-default-page.png" />

### Configure General Settings

2. By default, the system displays the **Settings** tab.

3. Enter a **Name** for the scorecard.

4. Enter a short optional **Description**.

5. Select the **Call Direction** to define applicable interactions—**Inbound**, **Outbound**, or **Both**.

6. Select the required **Languages** from the drop-down to load the associated evaluation metrics and attributes.

7. Select a **Not Applicable Scoring Treatment**:

   * **Exclude from Score** (default): Removes a Not Applicable metric's weightage from both the numerator and denominator when the system aggregates metric results into Agent Attribute and Scorecard scores.
   * **Award Full Weightage**: Adds a Not Applicable metric's weightage to both the numerator and denominator, so the metric contributes full marks to the Agent Attribute and Scorecard scores while its displayed outcome stays Not Applicable.

This setting uses a separate configuration from the **Not Applicable Scoring Treatment** configured on the evaluation form. It doesn't inherit the evaluation form's setting and applies only to future evaluations. When the two settings use different values, the scorecard score can differ from the corresponding evaluation form score.

7. Turn on to **Set the minimum duration required to complete evaluations**.

8. Enter a threshold value in minutes (**MIN**) and seconds (**SEC**).

9. Set the minimum **Pass Score** required for agents.

10. Select **Next**.

    <img src="https://mintcdn.com/koreai-content-gov/b5uJrPCb_sHn2XGc/ai-for-service/quality-ai/configure/agent-scorecards/images/add-new-agent-score-card-settings.png?fit=max&auto=format&n=b5uJrPCb_sHn2XGc&q=85&s=29bfb43e708395721f06f32a4867c65c" alt="Agent Scorecards Settings" width="495" height="625" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/add-new-agent-score-card-settings.png" />

<Note>At least one scorecard you must set as the default. The first scorecard you create becomes the default automatically. Before removing the default setting from a scorecard, assign another scorecard as the default.</Note>

### Configure Agent Attributes

Select and configure attributes that define agent performance scoring.

1. Search for the relevant agent attributes name.

2. Select **Agent Attributes** to add them to the scorecard.

   <img src="https://mintcdn.com/koreai-content-gov/uQy_3EJi2vwOWgyh/ai-for-service/quality-ai/configure/agent-scorecards/images/agent-attributes.png?fit=max&auto=format&n=uQy_3EJi2vwOWgyh&q=85&s=dfe95ed2fb39db14612afa447ca29d8b" alt="Add Agent Attributes" width="664" height="623" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/agent-attributes.png" />

   <Note>Only attributes matching the selected languages are shown. Each attribute contributes proportionally to the final score. The **Next** button is enabled only when total weightage equals 100%.</Note>

3. Select **Add Agent Attributes**.

4. Assign a **Weightage** to each selected attribute.

   <img src="https://mintcdn.com/koreai-content-gov/GTJFDEwn5iY7SphC/ai-for-service/quality-ai/configure/agent-scorecards/images/add-agent-attributes-weight.png?fit=max&auto=format&n=GTJFDEwn5iY7SphC&q=85&s=91541851dbf6f74faa181ed494368fde" alt="Add Agent Attributes Weight" width="552" height="936" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/add-agent-attributes-weight.png" />

5. Select **Next**.

   <Note>Only attributes matching the selected languages are shown. Each attribute contributes proportionally to the final score. The **Next** button is enabled only when total weightage equals 100%.</Note>

### Configure Assignments

Assign the scorecard to agents and agent groups.

1. Search and select agents or agent groups, and configure contact direction-based assignment.

2. Select **Add Agents** and **Add Groups** to include them in the scorecard.

   <img src="https://mintcdn.com/koreai-content-gov/qmsXYkCVLPlGCs-Y/ai-for-service/quality-ai/configure/agent-scorecards/images/add-agent-group.png?fit=max&auto=format&n=qmsXYkCVLPlGCs-Y&q=85&s=d7bda68002951f136998c748562a8d5a" alt="Add Agent Group" width="546" height="620" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/add-agent-group.png" />

3. Select **Create** to create the Agent Scorecard with the selected agent attributes and assigned agents and agent groups.

   <img src="https://mintcdn.com/koreai-content-gov/uQy_3EJi2vwOWgyh/ai-for-service/quality-ai/configure/agent-scorecards/images/newly-created-scorecard.png?fit=max&auto=format&n=uQy_3EJi2vwOWgyh&q=85&s=94633c6cc93475f169f7d594ddbbe74b" alt="New Scorecard" width="1348" height="557" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/newly-created-scorecard.png" />

   <Note>Toggle the Scorecard **Status** to **Active** to begin agent-level scoring.</Note>

### Agent Scorecard Elements

The Agent Scorecards elements list displays the following columns for each scorecard:

* **Name**: Name of the agent scorecard.
* **Agent Attribute**: List of performance attributes assigned to the scorecard.
* **Assignee**: Agents or agent groups assigned to the scorecard.
* **Created By**: Name of the user who created the scorecard.
* **Pass Score**: Minimum score threshold required for an agent to pass.
* **Status**: Indicates whether the scorecard is Active or Inactive.
* **Search**: Search for agent scorecards by name.

***

#### Not Applicable Scoring Treatment

The scorecard-level Not Applicable Scoring Treatment determines how Not Applicable metric outcomes contribute to Agent Attribute and overall Scorecard scores. The system applies this setting across the Agent Leaderboard, Agent Dashboard, and agent-specific dashboards. It doesn't inherit the **Not Applicable Scoring Treatment** configured on the evaluation form.

| **Treatment**                    | **Effect on Agent Attribute and Scorecard Scores**                                                                                                              |
| -------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Exclude from Score** (default) | Excludes the Not Applicable metric's weight from the score calculation by removing it from both the numerator and denominator.                                  |
| **Award Full Weightage**         | Includes the Not Applicable metric's weight as full credit by adding it to both the numerator and denominator. The metric remains marked as **Not Applicable**. |

A change to this setting doesn't retroactively rescore conversations evaluated. It applies only to metric contributions from conversations scored after the change.

<Note> Not Applicable\*\* scoring is configured independently for evaluation forms and scorecards. Different settings can result in different scores for the same interaction.</Note>

***

#### How Minimum Duration Threshold Works

The minimum duration threshold determines the shortest interaction that qualifies for AutoQA scoring.

| Contact Duration           | Status Assigned      | Result                                     |
| -------------------------- | -------------------- | ------------------------------------------ |
| Meets or exceeds threshold | —                    | Evaluated normally.                        |
| Falls below threshold      | Below Threshold      | Excluded from scoring and quality metrics. |
| Duration unresolved        | Duration unavailable | Excluded from evaluation.                  |

#### Duration Calculation by Channel

| Channel                  | Duration Measured As                                |
| ------------------------ | --------------------------------------------------- |
| Voice                    | Full call duration, including hold time.            |
| Chat                     | Time between the first and last message timestamps. |
| Quality AI Express (FTP) | Based on the `start_time` and `end_time` fields.    |

<Note>Threshold changes apply only to new contacts after the scorecard is saved. Disabling the threshold evaluates all contacts.</Note>

#### Manual Override

Supervisors with the appropriate permissions can manually evaluate contacts flagged as Below Threshold or Duration Unavailable. The system includes these manually evaluated contacts in quality metrics as normal

#### Versioning Behavior

The system versions threshold settings per scorecard. Updates apply only to contacts ingested after saving the scorecard. Historical contacts continue to use the previous threshold configuration.

#### Pass Score Logic

* **Pass**: Score ≥ Pass Score
* **Fail**: Score \< Pass Score.

Total attribute weight must equal 100%.

***

## Manage Agent Scorecards

### Edit an Agent Scorecard

1. Select the vertical ellipse (⋮) on any existing scorecard.

2. Select **Edit**.

3. Update the **Settings**, including **Not Applicable Scoring Treatment**, **Agent Attributes**, or **Assignments**, as needed.

4. Select **Update** to save.

***

#### Delete an Agent Scorecard

To delete an agent scorecard:

1. Select the vertical ellipsis (⋮) for the scorecard you want to delete.
2. Remove any unassigned attributes. If an attribute belongs to another scorecard, remove that scorecard assignment first.
3. Select **Delete** to permanently remove the scorecard.

<Note>Deleting an agent scorecard permanently removes all data associated with the scorecard. Deleted scorecards can't be recovered.</Note>

#### Delete the Default Scorecard

You can't delete the default scorecard directly.

Before deleting the current default scorecard:

1. Assign another scorecard as the default.
2. Select the vertical ellipsis (**⋮**) for the current default scorecard.
3. Select **Delete**.

If you attempt to delete the default scorecard without assigning another scorecard as the default, the system displays a warning dialog.

#### Delete an Agent Scorecard

1. Select the vertical ellipse (⋮) on any existing scorecard.

2. Delete attributes only if they're unassigned; otherwise, remove associated scorecards first.

3. Select **Delete** to permanently remove the selected agent scorecard.

#### Delete Agent Scorecard

When you delete an agent scorecard, you permanently lose all data linked to that scorecard, and you can't recover it.

#### Delete Default Scorecard

You can't delete the default scorecard directly. Before deleting any default scorecard, you must assign another scorecard as the default before deleting the current scorecard. Otherwise, the system displays a warning dialog.

### Language Configuration Warnings

#### Adding Languages

A warning appears if a new language isn't configured for existing metrics or attributes. You must configure metrics for that language first.

#### Deleting Languages

You can’t remove any languages linked to active scorecards or attributes until you clear those links. The system allows language modification only for attributes configured for the selected languages.

***

## Agent Attributes

This feature generates agent attributes by aggregating evaluation metrics to calculate an Agent Attribute Score, which highlights critical performance areas such as Professionalism and Product Knowledge. The system derives the score from evaluation metric data and filters the displayed attributes according to the scorecard’s language.

### Access Agent Attributes

Navigate to **Quality AI** > **CONFIGURE** > **Agent Attributes**.

<img src="https://mintcdn.com/koreai-content-gov/b5uJrPCb_sHn2XGc/ai-for-service/quality-ai/configure/agent-scorecards/images/agent-attributes-default-screen.png?fit=max&auto=format&n=b5uJrPCb_sHn2XGc&q=85&s=99e809e0a4d105ede838b3dc987d6c6f" alt="Agent Attributes" width="1361" height="631" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/agent-attributes-default-screen.png" />

### Create a New Agent Attribute

1. Select **Agent Attributes** > **+ New Agent Attribute**.

   <img src="https://mintcdn.com/koreai-content-gov/uQy_3EJi2vwOWgyh/ai-for-service/quality-ai/configure/agent-scorecards/images/new-agent-attribute1.png?fit=max&auto=format&n=uQy_3EJi2vwOWgyh&q=85&s=b7e3c59311ae644efb8674267850c3ca" alt="New Agent Attribute" width="1163" height="295" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/new-agent-attribute1.png" />

2. In the **Add New Agent Attribute** panel:

   * Enter a **Name**.
   * Enter an optional **Description**.
   * Select a target **Language** to load the associated evaluation metrics and attributes.
   * Search for and select evaluation metrics using **Add Evaluation Metrics**.

   <img src="https://mintcdn.com/koreai-content-gov/GTJFDEwn5iY7SphC/ai-for-service/quality-ai/configure/agent-scorecards/images/add-new-agent-attribute-details.png?fit=max&auto=format&n=GTJFDEwn5iY7SphC&q=85&s=f3294c2b1875f7e211059ec6a4eb1a2f" alt="Add New Agent Attribute" width="553" height="940" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/add-new-agent-attribute-details.png" />

#### Add Evaluation Metrics

1. Select **Add Evaluation Metric** to assign the selected metrics to the agent attribute.

2. Add **Total Positive Weightage** cumulative score percentage.

3. Add **Total Negative Weightage** cumulative score percentage.

#### Add Metric Configuration

1. Define the overall **Weightage** of the metric.
2. Assign a percentage value for each Outcome (for example, Yes or No).
3. Select a correct **Response** for the evaluated outcome.
4. Enable to mark the metric **Fatal Error** (if it's a critical failure condition).
5. Select **Create**.

   <img src="https://mintcdn.com/koreai-content-gov/GTJFDEwn5iY7SphC/ai-for-service/quality-ai/configure/agent-scorecards/images/add-agent-attributes-weight.png?fit=max&auto=format&n=GTJFDEwn5iY7SphC&q=85&s=91541851dbf6f74faa181ed494368fde" alt="Agent Attribute Weight" width="552" height="936" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/add-agent-attributes-weight.png" />

### Agent Attributes Elements

Each agent attribute displays the following elements list:

| Column          | Description                     |
| --------------- | ------------------------------- |
| **Name**        | Attribute name                  |
| **Description** | Optional description            |
| **Score Card**  | Status of all linked scorecards |

## Manage Agent Attributes

### Edit an Agent Attribute

1. Select the attribute and click the vertical ellipsis (⋮).

2. Select **Edit**.

3. Update the evaluation metrics, and select **Update**.

## Language Update Warnings

### Adding A New Language

A warning appears if the new language isn’t configured for the current by-question metrics. Configure the new language on the listed metrics or select a language that supports the metrics.

### Deleting A Language

The system displays a warning if the language is in use in other scorecards. Remove the language configuration from all associated scorecards before deleting it. The system blocks deletion if the language is still in use.

<Note>Only scorecards for the selected languages appear. You can remove or delete languages not linked to any scorecards.</Note>

### Delete an Agent Attribute

1. Select the three-dot (⋮) menu on the attribute you want to delete.

2. Select **Delete** and confirm the action in the warning dialog.

   <img src="https://mintcdn.com/koreai-content-gov/uQy_3EJi2vwOWgyh/ai-for-service/quality-ai/configure/agent-scorecards/images/delete-agent-attribute.png?fit=max&auto=format&n=uQy_3EJi2vwOWgyh&q=85&s=598a1170b49dd7b87e8540390c3a0f78" alt="Delete Attribute Warning" width="1158" height="196" data-path="ai-for-service/quality-ai/configure/agent-scorecards/images/delete-agent-attribute.png" />

### Agent Attribute Deletion Warnings

| Scenario                           | Behavior                                                                                  |
| ---------------------------------- | ----------------------------------------------------------------------------------------- |
| **Deleting any attribute**         | All data linked to the attribute is permanently deleted and can't be recovered            |
| **Deleting an assigned attribute** | You can’t delete an attribute assigned to a scorecard. Remove it from the scorecard first |
| **Eligible for deletion**          | Only unassigned agent attributes can be deleted.                                          |

<Note>You can't delete an attribute that's assigned to a scorecard. Remove it from all scorecards first. Deleting an attribute removes all associated data.</Note>

***
