⏱️ The Essentials in 3 Minutes |
🧠 Understand the Value of User Feedback
Feedback is not just a satisfaction indicator: it is an organizational learning signal. When a learner flags an incorrect or incomplete response, they are revealing a gap in your knowledge base or an ambiguity in your documents.
By reviewing these returns regularly, you transform your collaborators' day-to-day experience into a driver of training quality improvement. It is a concrete lever for aligning your knowledge base with your teams' actual needs - not what you assume they need.
💡 Access the Feedback Interface
Feedback is accessible in the Didask AI section > Feedback tab. This interface centralizes all ratings left by your users on AI responses.
For each feedback, you can see:
The precise date and time of the interaction
The rated AI response
The user's reaction (thumbs up or thumbs down)
Categorization tags (e.g. "Too verbose")
Optional comments left by the user
A Details button to access the full conversation context
You can filter feedback by period to analyze how quality evolves over time and identify trends or peaks in dissatisfaction.
🔍 Interpret Negative Feedback
Users can qualify their negative feedback with specific tags.
Content issues:
"Incorrect response": incorrect or outdated information in your knowledge base
"Incomplete response": missing information on the requested topic
"Off-topic": the AI did not understand the question or drifted to another subject
Format issues:
"Too verbose": responses too long, lacking synthesis
"Too brief": responses insufficiently detailed
"Instructions not followed": the AI did not follow the user's specific instructions
By clicking "Details", you access the full conversation context, the response generation time, the sources used by the AI, and the exchange history.
✏️ Act on Your Knowledge Base in Response to Feedback
When facing recurring negative feedback:
Add content on poorly covered topics
Update the outdated information identified
Remove contradictory sources
Reorganize documents for better readability and relevance
🤝 Escalate to the Didask Team If Needed
Feedback is automatically aggregated on the Didask side. The technical team can identify and correct systemic issues. Contact support for recurring problems that cannot be resolved by improving your content.
✨ Optimization Tips
Responsiveness: review feedback weekly to quickly identify issues.
Communication: inform your users of improvements made following their feedback.
Qualitative analysis: beyond volumes, read the comments to understand actual expectations.
Post-correction monitoring: verify that improvements effectively reduce negative feedback on the same topics.
Keywords: feedback, response quality, Didask AI, coach, learning assistant, monitoring, continuous improvement.