An explained result
A reason accompanies the score for review.
Review quality, not just response time
AI scores closed conversations and explains its result. Compare its findings with customer feedback and a human review to identify specific improvements.
For your team
Manager involvement
A reviewer adds their own score and comment.
Agent feedback
Feedback relates to a specific closed conversation.
What Tygy can do
After resolution
Start with the reason behind the score
After an eligible conversation closes, AI analyses the messages using the quality methodology. The score and explanation help your team choose cases to review.
- Review a completed conversation rather than a single reply.
- Read the explanation alongside the original conversation.
- Keep in mind that not every conversation is eligible.
Human review
Three perspectives give you more context
AI, the customer and a reviewer see the outcome from different perspectives. Compare available scores and leave the agent a comment with a specific example.
- See customer feedback when a rating was submitted.
- Add your own score and explanation.
- The agent can acknowledge the feedback.
Team standards
Define what a good answer means
Configure the quality methodology for your team’s work. Reviewer corrections provide examples for subsequent AI evaluations.
- Make your expectations for replies explicit.
- Review differences between AI and human scores.
- Turn useful findings into team guidance.
Screenshots show the Tygy interface with demonstration data. Names, conversations and figures are examples.
How it works
From a closed conversation to an improvement
A score is useful when it leads to a clear action.
Close
The team closes the conversation.
Evaluate
AI analyses an eligible conversation.
Review
A reviewer compares the findings with the context.
Discuss
The agent receives specific feedback.
Configure the feature for your team and check the results against your own examples.
Everyday tools
Make quality review a routine
Methodology
Set criteria that match your support work.
Reviewer permissions
User permissions determine access to reviews.
Outcome context
Consider customer feedback and the AI explanation alongside the messages.
AI quality
Questions and answers
Does AI score each message?
No. A score applies to a closed conversation. Evaluation requires an eligible two-way exchange.
Can a reviewer disagree with AI?
Yes. A reviewer can add their own score and explanation, shown separately from the AI score.
Why might a score be missing?
The conversation may be ineligible, the feature may be disabled, or evaluation may still be processing. AI also needs an available token balance.
Start with a few real conversations
Compare AI and reviewer scores, refine your criteria and make reviews part of your team’s work.
14 days free · No credit card required