Is there a way to collect or access thumbs-up or down feedbacks for responses in the custom chat agent or an action I created? I wonder if I can stream those to analytics layer, directly access from UI or need any onboarding steps to access feedback data for chat agents or actions.
3 Likes
- Are you looking to access feedback data directly through the platform’s UI, or would an API or event stream integration work better for your use case?
- Is your goal to use the feedback for model improvement, analytics, or user experience reporting?
- I expected the same UI like Q Topic for tracking feedbacks. API or event streaming works too.
- We need it mainly for model improvement based on user experience reporting. I would like to know who used the agent/action and how they interacted with, which one they were satisfied or not satisfied, or capture any inaccurate data from the chat answer they report.
@priyaik Could you share if feedback data access is available?
Yes, you can collect a feedback from a customer chat agent or an MCP server action in Amazon Quick-Sight by enabling a feedback mechanism within your setup. This typically involves capturing user responses (like thumbs-up/down or text feedback) and storing them in a data source such as Amazon S3 or DynamoDB.
Once stored, you can connect the data to Quick-Sight to analyse the patterns, satisfaction levels, or agent performance. Make sure your agent or back-end action is configured to send feedback data consistently and that permissions are properly set.
After that, you can easily visualise feedback trends or create dashboards in Quick-Sight to understand how users interact with your chat agent and where improvements can be made.
@Ahmad-Talha Can you share any reference or documentation to enable feedback mechanism from QuickSight UI?
Hi @hoyeon — thanks for following up. 
Here are a couple of reference / documentation links regarding feedback mechanisms and how they (and related features) work in the Amazon QuickSight / Amazon Q Business ecosystem:
-
There’s a page on “User feedback using Amazon Q Business APIs” which explains how end-user feedback is captured in chat/assistant workflows (via
PutFeedback). AWS Documentation+1 -
Within QuickSight topics (for natural-language Q), you can review user feedback and usage in the “Summary” and “User Activity” tabs. Amazon Web Services, Inc.+1
-
A community discussion indicates that embedding a generic feedback prompt (e.g., on dashboard close or download) is not supported out-of-the-box in QuickSight, but can be implemented by embedding the dashboard in your own UI and then layering custom feedback components. Amazon Quick Community+1
So to answer your question:
You can capture feedback (especially for chat/assistant flows) via Amazon Q Business API. For standard QuickSight dashboards, native “thumbs up/down” prompts are not built-in, but you can integrate feedback by embedding QuickSight in your app/portal and hooking up custom feedback widgets.
If you like, I can dig up a full step-by-step guide (with sample code) for how to inject a feedback form into a QuickSight dashboard embed and then pull that data into QuickSight for analysis. Would that be helpful?
2 Likes
Hi @Ahmad-Talha ,
Appreciate for your detailed explanation.
The full step-by-step guide with sample code will be definitely helpful for me.
I do not need Q Business API unless the same API would work with Quick action or custom agent.
I know QuickSight topic has the user activity tab for feedback and usage. I noticed that Quick action has usage tab but there was no data yet even when there are active users.
Having QuickSight dashboard for the Quick Action feedback is intriguing. I wonder if the linked Q&A would work the same as it was posted 2 years ago. I do want to know step-by-step for Quick agent or action feedback data gathering (Not Q topic, Not Q business) if it works.
Thanks for the detailed clarification, @Hoyeon! That’s a great point — the Quick Action feedback tracking isn’t as well-documented as Q topics. I’ll look into how feedback data can be gathered specifically for Quick agents and actions, separate from Q Business. Once confirmed, I’ll share a clear step-by-step guide with example queries or API calls. Your observation about the missing usage data is really insightful!
1 Like
@Ahmad-Talha This feature is critical for my project. I’ll wait for your step-by-step guide with example.
Hi @Hoyeon,
Thanks for confirming — that really helps narrow the focus. I’ve started exploring how feedback and usage tracking work specifically for Quick actions and agents, separate from Q Business and Q Topics.
From what I’ve seen so far, the feedback data may not be directly exposed yet through the same APIs, but there are a few possible workarounds using CloudWatch metrics and QuickSight dataset integration. I’ll validate the approach and share a complete step-by-step guide with example queries once confirmed.
Appreciate your patience — I’ll update you soon with verified details.
@Ahmad-Talha Just checking in if you find any solution for this.
Hi @Ahmad-Talha,
Hope everything is well! Just checking in here to see if you are still able to share your feedback for @hoyeon’s initial question.
If there is no further message within 3 business days, I’ll go ahead and close out of this topic.
Thank you!
@priyaik Can you answer this question?
@hoyeon You can use Amazon CloudWatch Logs to deliver user conversations and response feedback in Amazon Quick for you to analyze. These logs can be delivered to multiple destinations, such as CloudWatch, Amazon S3, or (standard rates apply). We recommend that you set up conversation and feedback logging shortly after creating your Amazon Quick chat agent.
Hi, @hoyeon Did this solution work for you? I am marking this reply as, “Solution,” but let us know if this is not resolved. Thanks for posting your questions on the Quick Community Q&A Forum!