Skip to main content

Performing Detailed Interaction Analysis in Vela

The Interactions section is your comprehensive workspace for Quality Assurance (QA). Here, you can move beyond performance scores and dive deep into the specific details of a customer conversation using Vela's AI insights and manual review tools.

1. Accessing and Filtering Interactions​

To perform a detailed analysis, you first need to find the specific interaction that requires your attention.

A. Locate the Interaction​

  1. Navigate to the Interactions section (Calls or Chats tab).
  2. Use the Search through interaction transcript bar to find a conversation by keywords or phrases.
  3. Apply Advanced Filters to narrow down your selection:
    • Date Range: Focus on a specific time period.
    • Agent/Team: Filter by a specific agent you're coaching.
    • Agent Score: Filter for interactions with low automatic scores.
    • Tags: Use classification labels (e.g., "complaint", "sales") to group related calls.

B. View Interaction Details​

The list view provides critical metadata that helps you prioritise the review:

  • Handle and silent time: Review the duration of the conversation and any significant silent gaps.
  • Agent Score/Initial Score: See the score automatically assigned by Vela's AI.
  • Customer Sentiment: Quick overview of the customer's overall mood (Positive, Neutral, Negative).
  • Number of Alerts: Quickly identify calls that triggered a Smart Search rule.

2. Analysing the Interaction Transcript and Audio​

The Interaction Detail View combines the full transcript with the audio/chat content, synchronised with all AI analysis.

A. Utilise Transcription and Playback​

  1. Click on the interaction to open the Detail View.
  2. Listen to the call recording or read the chat transcript.
  3. Use the playback Speed adjustment (e.g., 1.25x or 1.5x) for efficient review.
  4. The Transcription is automatically processed in multiple languages (including African languages).
  5. Click on a Timestamp in the transcript to jump to the exact moment in the audio.

B. Review Smart Detector Insights​

The AI analysis provides context and flags specific moments for you:

InsightPurposeHow to Use in Review
SummaryAI-generated interaction overview.Get a quick sense of the key issue and resolution before diving into detail.
SentimentTracks the customer's emotional state throughout the conversation.Pay attention to moments where sentiment shifts from Positive to Negative, and check the agent's response at that time.
Pain PointsAI-identified customer frustration indicators.See if the agent successfully addressed or alleviated the pain points identified.
Keywords & IntentsImportant terms mentioned and the customer's goal (e.g., Sales, Complaint).Verify if the agent used required scripting (Keywords) and effectively handled the customer's primary objective (Intent).

3. Applying Quality Judgment and Feedback​

A detailed analysis culminates in either validating the AI's score or applying your manual judgment and providing actionable coaching feedback.

A. Complete the Scorecard​

  1. Review the Automatic Scorecard provided by Vela AI.
  2. If you agree, you can finish the review process. If not, click to "Create Manual Scorecard".
  3. Use the Manual Scorecard to apply your expertise and context that the AI may have missed.
  4. The final score from the manual scorecard will override the automatic score for performance tracking.

B. Collaborate and Coach​

Use the built-in communication tools to give timely feedback directly related to the interaction.

  1. Use the Comment System within the interaction detail view.
  2. Add specific comments or notes to the interaction transcript at relevant timestamps.
  3. Send and view comments for collaboration with other Team Leads or Managers.
  4. Send direct feedback to the agent, who will receive it in their dedicated Agent Portal notifications.

The expected outcome of a detailed analysis is a thorough quality assessment of the customer interaction with documented findings and actionable feedback.