-
Ledger
-
Test execution type
- Automated integration testing
- Automated UI testing
- Manual testing
-
Test execution status
- Pass
-
Fail
- Fail
- Re test
- Pending
-
Design Changes
-
Verify the Color changes For
-
Evaluation Header
- Pass
-
Comment Box
- Pass
-
Feedback Box
- Pass
-
Evaluation Header Verfication
-
Pass
- Pass
-
Remove the scroll bar
-
Verify the scrollbar should only be visible when user is hovering on evaluation screen
- Re test
-
Verify the above flow from other screen
- Re test
-
Change in audio log colour
-
Verify the colour of the audio log
- Pass
-
Verify the following components as per the new design
-
Description
- Pass
-
Feedback
-
Adding a feedback designs
- Verify from design
- Pass
- save button should be disable when there is no changes
- Pass
-
Editing a feedback design
- Verify from design
- Pass
- save button should be disable when there is no changes
- Pass
-
History feedback icon
- Pass
-
Original feedback text box
- Thumbs up and thumbs down button
- Pass
-
No feedback design
- Pass
- Pass
-
Feedback draft state
- Pass
- Pass
-
Feedback on emails
- Pass
-
Thumbs up design
- Should show only after mouse hover
- Pass
- Pass
- Same after evaluation
-
Thumb down design
- Pass
-
Thumb down text box design
- Pass
-
Comment
- Pass
-
Audio buttons
-
Verify the timestamp
-
Call
- Pass
-
chat
- Pass
-
email
- Pass
- Subtopic 7
-
Functional Flows
-
Ability to Provide Feedback
-
Thumbs Up/Down Functionality
-
Verify users can give thumbs up on auto-QA’ed questions.
- Pass
-
Verify users can give thumbs down on auto-QA’ed questions.
- Pass
-
Login with different users for same conversation and verify the thumbs up and down flow
- Pass
-
Additional Feedback on Thumbs Down
-
Verify users can provide additional feedback when giving thumbs down on auto-QA’ed questions.
- Pass
-
Verify users can select "Evidence drafted by Auto-QA is incorrect" when providing additional feedback on thumbs down.
- Pass
-
Verify users can select "Answer selected by Auto-QA is incorrect" when providing additional feedback on thumbs down.
- Pass
- Pass
-
Verify users can enter additional details in the text field when providing feedback on thumbs down.
- Pass
- Pass
-
Verify the feedback stored for each user in meta base
- Re test
-
Unified Evidence Experience
-
Metric Tag Evidence
-
Verify the presentation of metric tag evidence adheres to the specified format.
- When metric tag threshold was followed
- *Name of the metric tag*in this conversation was *below/above sort of operator* *duration*
- Pass
- When metric tag threshold was not followed
- *Name of the metric tag* in this conversation was not *below/above sort of operator* *duration*
- Pass
- When metric tag is n/a
-
Conversation Tag Evidence
-
Verify the presentation of conversation tag evidence aligns with the defined structure.
- When conversation tag was followed (Agent tags)
- The agent followed *Name of the conversation tag* in this conversation. The agent said:
- Pass
- *Timestamp* : *Utterances that we show in snippets
- Pass
- When conversation tag was not followed (Agent tags)
- The agent did not follow *Name of the conversation tag* in this conversation. The agent said:
*Timestamp* : *Utterances that we show in snippets*
- Re test
- When conversation tag was followed (Dynamic tags) For ex Convo ID 91237
- The agent followed *Name of the conversation tag* in this conversation. The agent said:
*Timestamp* : *Customer utterances that we show in snippets*
*Timestamp* : *Agent utterances that we show in snippets*
- Re test
-
When conversation tag was not followed (Agent tags) For ex Convo ID 91251
- The agent did not follow *Name of the conversation tag* in this conversation when the customer said:
*Timestamp* : *Customer utterances that we show in snippets*
- Pass
-
When conversation tag doesn’t exist (Dynamic tags without customer utterances) For ex, convo ID 91238
- The customer did not mention phrases or words related to *name of the conversation tag*.
- Pass
-
Guidelines Based Evidence Format
-
Verify autoqa guideline is working
- Pass
-
Verify the breakdown of guidelines evidence in the specified format.
- Guidelines evidence should be broken down into this format :
*Summary*. The agent said :
*Timestamp* *Utterances*
*Timestamp* *Utterances*
*Timestamp* *Utterances*
- Pass
-
Timestamp Experience
-
Verify clicking on play button or timestamp directs the user to the exact utterance in the transcript or call map.
- Call
- The exact utterance in the transcript (call, chat and email)
- Pass
- The exact moment in the call map (in case of call)
- Pass
- Verify clicking on a timestamp in call evidence takes the user to the specified time in the audio timeline and highlights the full utterance.
- Eg: Start time:
- Clicking on timestamp (such as 01:35 - highlighted in yellow) takes the user to 01:35 min in the audio timeline, plays the audio and highlights the full utterance (yellow as per the evidence) or exact words in parallel (whichever requires less effort for V1)
- Pass
- Verify the audio continues playing till the end of the highlighted utterance.
- End time
- The audio should continue playing till the end of the utterance (end of yellow part)
- Pass
- Chat
- The exact utterance in the transcript (call, chat and email)
- Verify clicking on a timestamp in chat & email evidence highlights the exact utterance in the transcript.
- Verify the entire utterance presented in the evidence is highlighted.
- Start time:
- Clicking on timestamp (such as 01:35) highlights the exact utterance in the transcript (yellow as per the evidence)
End time
- The whole utterance (as per specific timestamp) presented in the evidence should be highlighted
- Pass
- Email
- The exact utterance in the transcript (call, chat and email)
- Verify clicking on a timestamp in chat & email evidence highlights the exact utterance in the transcript.
- Verify the entire utterance presented in the evidence is highlighted.
- Verify how transcript identification works for email conversations when using timestamps.
- Start time:
- Clicking on timestamp (such as 01:35) highlights the exact utterance in the transcript (yellow as per the evidence)
End time
- The whole utterance (as per specific timestamp) presented in the evidence should be highlighted
- Pass
-
Edit of time stamp
- Verify that users can edit timestamps within the specified format (e.g., 00:00).
- Pass
- Verify that editing timestamps in the middle of the evidence line does not affect the timestamp format.
- Pass
-
Metabase verification
-
Verify the following data are stored in metabase
- Tenant/Org
- Pass
- Question
- Pass
- Convo ID
- Pass
- Pass
- Original Answer (auto-QA’ed)
- Pass
- Updated Answer
- Pass
- Date & time of changes
- Pass
- Verify that the Metabase table captures changes in custom form answers and evidences, and confirm the workflow from suggestion to final suggestion to user-edited
-
Draft comment flows
- . Verify if users can save feedback drafts before submitting the edited feedback.
- Verify the state management for draft feedback, ensuring the last feedback is retained for the submitted evaluation.
- Verify other draft feedback flows
-
Add/ Edit Evidence Experience
-
Edit Auto-QA’ed Evidences
-
Verify users can edit auto-QA’ed evidences using a rich text editor.
- Pass
-
Verify users can rephrase words/phrases, modify timestamps, and save changes.
- Pass
-
Verify the flow after clearing the evidence, and saving it,
- Pass
-
User Detail on Manual Evidence Edit
-
Verify the title changes to “Evidence - [name of the user]” when evidence is manually edited.
- Pass
-
Add Evidence for Manual Question
-
Verify users can add evidence to manual questions in the manual rubric, including timestamp conversion.
- Pass
-
Verify users can go to the exact place in the transcript and call map using added timestamp.
- Pass
-
Overwriting of Feedback
-
. Verify that if multiple users provide feedback, the final version, based on permissions, supersedes.
- Verify the permissions chart for editing and confirm that the final version is shown before submission.
- Pass
-
Multiple Feedback
-
Verify that multiple feedback entries are not allowed for a single evaluation.
- Pass
-
Comment Button Visiblity
-
Verify the comment button is enabled only after evaluation submission.
- Pass
-
Verify all the previous comment flows
- Pass
-
Verify the accept and dispute flows for comments
- Pass
-
Export Feature
- Verify the Evaluation Comments should be downloaded in CSV
- Verify the Evaluation Feedback should be downloaded separately in CSV
- Verify the export file
- Check with plat team for any export functionality issues
-
Calibration Flow Changes
-
Calibration Review Screen Changes
-
Verify all users can view, edit, and add AI responses and evidences in the calibration review screen, with applicable UI changes.
- View
- Pass
- Edit
- Pass
- Add
- Pass
-
Verify remaining same flow as evaluation
- Pass
-
Verify that the new UI elements (e.g., header) are visible in the calibration review screen.
- Pass
-
: Calibration Report View Changes
-
Verify only user-edited evidences and manual evidences are visible in the calibration report view.
- Pass
-
Verify that comments are replaced by user-edited evidences in the calibration report view.
- Pass
-
Transcript/Snippet Highlighting
-
Verify that when switching between transcript/snippet tabs, evidence for conversation tags is always directed to the transcript.
-
Verify that the connectivity is shown even when users switch between tabs.
- Pass
-
Verify the handling of metric tag "dead air" where it points to the snippet with no connectivity to transcript.
- Pass
-
Verify that when data is not available, it connects to the snippet.
- Pass
-
Post-Evaluation Actions
-
Verify if users can provide thumbs up or thumbs down after the evaluation
-
Pass
- Pass
-
Verify if evidence editing is allowed post-evaluation.
- Pass
-
Edge cases
-
Evaluation delete
-
In case of evaluation deletion, manually added feedback/evidence will keep back or will delete? We need to retain the auto-QA feedback and keep the rest of the behaviour same as comment
- Pass
-
Page refresh
-
if a user does thumbs up or thumbs down and he refreshes the page then on hover blue thumbs up/down should be visible or white thumbs up/down
- Pass
-
Thumbs down are independent of users
-
Login with different users and verfy
- Pass
-
Different Versions of Edited Evidence
-
Confirm that different users can have different versions of edited evidence and verify the support for drafts.
- Pass
-
Rubric Changes
-
Verify that changes in evidence remain safe and visible when users choose back the same rubric after a rubric change.
- Pass
-
Audit Log, Notification & Email related changes
-
Verify the regression for these flows
- Pass
-
Permission cases for feedback
-
Adding Evidence
-
Verify QA Auditor can add evidence.
- Pass
-
Verify Admins/SA can add evidence.
- Pass
-
Verify Agent cannot add evidence.
- Pass
-
Verify Manager cannot add evidence.
- Pass
-
Editing Evidence
-
Verify QA Auditor can edit their team's evidence.
- Fail
-
Verify Admins/SA can edit any evidence.
- Pass
-
Verify Agent cannot edit evidence.
- Pass
-
Verify Manager cannot edit evidence
- Pass
-
Feedback Provision
-
Verify Agent cannot provide feedback.
- Pass
-
Verify Manager cannot provide feedback.
- Pass
-
Verify QA Auditor can provide feedback on evidence.
- Pass
-
Verify Admins/SA can provide feedback on any evidence.
- Pass
-
Evidence Review
-
Verify Agent can view edited evidence.
- Pass
-
Verify Manager can view edited evidence.
- Pass
-
Verify QA Auditor can view edited evidence.
- Pass
-
Verify Admins/SA can view edited evidence.
- Pass
-
Original Evidence Visibility
-
Verify Agent cannot view original evidence once edited.
- Pass
-
Verify Manager cannot view original evidence once edited.
- Pass
-
Verify QA Auditor can view original evidence once edited.
- Pass
-
Verify Admins/SA can view original evidence once edited.
- Pass
-
Access to Unedited Evidence
-
Verify Agent can view original unedited evidence.
- Pass
-
Verify Manager can view original unedited evidence.
- Pass
-
Verify QA Auditor can view original unedited evidence.
- Pass
-
Verify Admins/SA can view original unedited evidence.
- Pass
-
Extras
-
States for user when they dont have permission to evaluate
-
Dont show empty feedback boxes
- Pass
-
Dont show the lower state of the feedback boxes
- Pass
-
Multiroles
-
Verify that users with multi-role permissions have the expected access.
- Pending