Skip to content

Conversations

The Conversations tab is your command center for managing, reviewing, and analyzing call recordings in Voice QA.


Overview

All call recordings flow into the Conversations queue where they're:

  • Transcribed to searchable text
  • Evaluated against your rules
  • Scored and analyzed
  • Available for review

The Conversation Queue

What You See

Each conversation in the list shows:

ColumnDescription
Date/TimeWhen the call was recorded
DurationLength of the recording
AgentWho handled the call
ClientCustomer (if identified)
StatusProcessing state
ScoreEvaluation result (if complete)
SentimentPositive, neutral, or negative

Processing Status

Conversations move through these stages:

StatusDescription
PendingUploaded, waiting to process
TranscribingConverting speech to text
TranscribedText ready, awaiting evaluation
ProcessingAI evaluation in progress
CompletedFully evaluated with score
FailedError occurred (can retry)

Uploading Conversations

Manual Upload

  1. Click Upload Conversation button
  2. In the dialog:
    • Select Department
    • Select Agent (or leave unknown)
    • Choose audio file(s)
  3. Click Upload
  4. Wait for processing

Supported Formats

FormatExtension
MP3.mp3
WAV.wav
M4A.m4a
OGG.ogg
WebM.webm

Maximum file size: 50MB per file

Automatic Upload

With integrations configured, calls upload automatically:

  1. Call ends in your phone system
  2. Recording sent to Voice QA
  3. Transcription begins immediately
  4. Evaluation runs automatically
  5. Results appear in dashboard

Learn more about Integration →


Filtering Conversations

Find specific calls using the filter bar:

Available Filters

FilterOptions
Date RangeToday, Last 7 days, Last 30 days, Custom
DepartmentSelect specific team
AgentFilter by staff member
StatusPending, Processing, Completed, Failed
Score RangeMin/max score percentage
SentimentPositive, Neutral, Negative
Call TypeIncoming, Outgoing

Combining Filters

Filters work together. For example:

  • "Show all negative-sentiment calls from Sales in the past week"
  • "Show completed evaluations for Agent John scoring below 70%"
  • "Show all pending calls from Customer Support"

Conversation Details

Click any conversation to see its full evaluation.

Audio Player

  • Play/pause the recording
  • Skip forward/backward
  • Adjust playback speed
  • Jump to specific timestamps

Call Information

FieldDescription
Recording DateWhen the call occurred
DurationLength of the call
AgentWho handled it
ClientCustomer (if identified)
Credits UsedProcessing cost

Evaluation Report

Overall Assessment:

  • Score — Percentage based on rules
  • Sentiment — Customer emotional tone
  • Satisfaction — Predicted customer satisfaction
  • Summary — AI-generated call summary
  • Recommendations — Improvement suggestions

Rule-by-Rule Breakdown:

For each evaluation criterion:

FieldDescription
ScorePoints earned vs. maximum
EvidenceQuotes from the conversation
ReasoningWhy this score was given
SuggestionsHow to improve

Transcript

Full conversation text with:

  • Speaker labels (Agent/Customer)
  • Timestamps
  • Searchable content

Manual Actions

Re-Transcribe

If transcription quality is poor:

  1. Open the conversation
  2. Click Re-transcribe
  3. Wait for new transcription

Re-Evaluate

If rules have changed:

  1. Open the conversation
  2. Click Re-evaluate
  3. AI applies current rules

Delete

Remove a conversation:

  1. Open the conversation
  2. Click Delete
  3. Confirm deletion

⚠️ Deletion is permanent and cannot be undone.


Bulk Actions

Select multiple conversations to:

  • Delete in bulk
  • Re-evaluate with updated rules
  • Export data

Keyboard Shortcuts

Speed up evaluation review:

KeyAction
SpacePlay/Pause audio
Skip back/forward 5 seconds
Previous/Next rule

Tips for Reviewing

Efficient Review Process

  1. Use filters to focus on specific needs
  2. Sort by score to find outliers
  3. Review failed rules for patterns
  4. Use keyboard shortcuts for speed

What to Look For

  • Low scores — Identify training needs
  • Negative sentiment — At-risk customers
  • Common failures — Rule or process issues
  • High performers — Best practices to share

Next Steps