Who can access ?
Roles | Access |
👑 Admins | Full access |
🛠️ Editors | No access |
👀 Viewers | No access |
🔒 Restricted Viewers | No access |
Where can I get the API access of my platform ?
Organization settings -> Advanced settings
Public key | Private key |
A non-sensitive identifier for your organization (safe to expose, but currently not used for auth). | Your secret API access token (keep it safe, never expose it publicly).
This is the key you actually use to authenticate API calls to the Feedier Public API (endpoints under /v3).
It works as a Bearer token: |
When to use the Webhook in Organization settings ?
The organization settings Webhook allows you to listen the event : New user joined Feedier to make an action out of it.
What is the noise detection ?
Noise detection is an automatic quality filter for verbatims. Think of it like a spam filter for your emails — but instead of filtering junk mail, it filters out low-quality or meaningless feedback that would pollute your analysis.
How does it work?
When someone submits a text response to a survey, the system automatically reads it using AI.
The AI decides: "Is this actual meaningful feedback, or is it noise?" (e.g. random characters, gibberish, test entries, irrelevant text).
If it's noise → the response is flagged and excluded from your reports, topics, and sentiment analysis.
If it's legitimate → the response proceeds normally through analysis (sentiment, topics, satisfaction scores, etc.).
💡 When exporting a list of Feedback if a question does not appear in your export it might be because the corresponding verbatim answers are considered as noise. You can export specifically noisy Feedbacks.
What happens to "Noisy" responses ?
They are not deleted, they're just set aside.
They don't pollute your dashboards, reports, or satisfaction scores
You can still see them and manually override the AI's decision if you disagree
You can export all noise-flagged verbatims into an Excel file for review
Who controls it?It can be turned on or off per organization in Settings → Advanced → Noise Detection
When you turn it off, all previously flagged responses are automatically unmarked and reprocessed into your analysis
Any user can manually mark or unmark a specific response as noise.