Introduction

The "3. Evaluation Detail by Attribute" report, available in the 'Reports' > 'Quality' section of wolkvox Manager, allows you to review the result of each evaluation at the individual attribute level. Unlike consolidated reports, this report shows the detail of each criterion evaluated within a conversation, indicating whether the agent met the attribute, if that attribute corresponds to a critical or non-critical error, and what comment was associated with the evaluation.

This report is useful for auditing quality evaluations in greater depth, understanding exactly where an agent failed or succeeded, and analyzing the details of the scoring in both manual evaluations and evaluations performed automatically with AutoQAi.

It is important to note that:

• If the evaluation was performed by a human quality analyst, the matrix, attributes, and error classification correspond to a manual configuration;
• If the evaluation was performed by AutoQAi, the matrix, attributes, and criteria are generated and analyzed automatically by artificial intelligence.

Report Information

The columns in this report include the following information:

• CONN_ID: Identification number of the evaluated conversation.
• AGENT_ID: Agent's extension number within the wolkvox system.
• AGENT_NAME: Name of the evaluated agent.
• ATTRIBUTE: Specific attribute or criterion of the quality matrix that was evaluated within the conversation.
• TYPE_ERROR: Type of error associated with the evaluated attribute. This column can show the following values:
  • CE: Critical error.
  • NCE: Non-critical error.
• YES/NO: Indicates whether the agent met the evaluated attribute. This column can show the following values:
  • yes: The agent met the attribute.
  • no: The agent did not meet the attribute.
• DATE: Date and time when the evaluation or scoring was performed.
• SURVEY: Name of the quality matrix with which the agent's interaction was scored.
  • When the matrix name includes the prefix AutoQAi_, it indicates that the evaluation was performed automatically with artificial intelligence.
• CATEGORY: Category to which the evaluated attribute belongs within the quality matrix. This column allows grouping criteria by topics or evaluation blocks, such as closing, solution, technical skills, case management, or communication.
• QUALITY_ANALYST: Name of the quality analyst who performed the evaluation. This column can show:
  • QAi_Bot: Indicates that the evaluation was performed automatically by AutoQAi.
  • The name of a quality analyst: Indicates that the evaluation was performed manually by a person.
• CHANNEL: Contact channel of the evaluated conversation. This column can show one of the following values:
  • chat: Chat interaction.
  • chat-web: Chat interaction from the web widget.
  • chat-whatsapp: Chat interaction from WhatsApp.
  • voice: Voice call.
• AI_COMMENT: Comment associated with the attribute evaluation. Although the column name refers to artificial intelligence, this field is used to show the observation or justification left by whoever performed the evaluation, whether:
  • The AutoQAi automatic system.
  • A human quality analyst, when applicable.
    • In practice, this column helps understand why the attribute was marked as met or unmet.

Important Consideration About This Report

This report shows one row per evaluated attribute, not one row per complete evaluation. This means that the same conversation can appear multiple times, once for each criterion reviewed in the quality matrix. Therefore, this report is especially useful when you need to review in detail the agent's behavior for each evaluated point and not just the final consolidated result.