Feedback
Last updated
Last updated
To access feedback data, navigate to the "Feedbacks" section within the Instinct AI dashboard. You can download the data in a CSV format for further analysis.
The dashboard offers several key metrics to help you understand feedback trends:
Confidence Level: The AI's confidence in providing accurate and relevant answers.
Unconfidence Level: The AI has a moderate level of confidence in its response
Helpful: User perception of the answer's usefulness.
Not Helpful: User perception of the answer's being not useful
The feedback data includes the following columns:
Question: The user's original query.
Answer: The AI's response to the query.
Citation: The source of information used by the AI (if applicable).
Confidence: The AI's confidence level in the response.
Answer Type: The type of response generated by the AI (e.g., informative, conversational, transactional).
Agent Name: The AI model or agent responsible for the response.
Helpful: Indicates whether the user found the response helpful.
Not Helpful: Indicates whether the user found the response not helpful.
Action: Any actions taken on the feedback (e.g., resolved, escalated).
By actively incorporating feedback into our AI development process, we achieve several key benefits:
Continuously Improving Accuracy: Our AI is constantly learning and evolving, leading to more accurate and informative responses over time.
Enhanced User Experience: By addressing user feedback and pain points, we create a more intuitive and satisfying AI experience for everyone.
