Insights
The Insights view (/staff/insights/:roomName) is the primary place to review, analyze, and act on all the feedback you collect, both the quick emoji-and-comment feedback from a room and the responses to your forms.
Navigating to Insights
- Click the chart icon on any room card in the Create view
- Click Responses inside a room's settings panel
- Navigate directly via URL:
/staff/insights/your-room-name
Global Filters
All tools in the Insights view share a single set of filters. Setting a filter applies it across every tool — you don't need to re-filter when switching between tabs.
| Filter | Options | Notes |
|---|---|---|
| Sentiment | All / Positive / Neutral / Negative | Narrows to responses matching that sentiment |
| Date range | Today / Last week / Last month / Last 6 months / Last year / All time / Custom | Custom lets you pick exact start and end dates |
| Event type | All / specific event types | Filters by the session or event the feedback was collected for |
| Search | Free text | Searches across comment text; debounced |
Filters reset automatically when switching between the feedback tools and the forms tools (e.g. going from Feedback list to Form cloud), except for the date range which persists.
You can deep-link into a filtered view using URL query params: ?sentiment=negative&days=30.
Tools
Use the tab bar to switch between the following tools.
Feedback list
Every individual response listed with:
- Emoji reaction (happy / neutral / sad)
- Written comment
- AI-generated sentiment label
- Timestamp and event metadata
Supports all global filters plus a text search across comment text.
Feedback trends
A time-series chart showing response volume and sentiment distribution over time. Useful for spotting spikes in negative feedback or drops in engagement after an event.
Word cloud
A visual word cloud of the most frequent terms across all responses. Larger words appear more often. You can exclude specific words — exclusions are saved per-browser.
Serious concerns
Responses automatically flagged by the AI as requiring immediate human review — covering harm, self-harm, threats, harassment, abuse, or severe misconduct.
Each concern shows:
- The original feedback text
- The AI's reasoning for flagging it
- Event type and instructor ID (if present)
- Jira ticket link (if your org has Jira connected and a ticket was created)
- Who resolved it and when, plus any resolution note
Resolve / Reopen workflow: Click Resolve on any concern to mark it handled. Optionally add a resolution note for your records. Resolved concerns can be reopened. Use the Unresolved / Resolved / All filter to switch views — the default is unresolved so nothing slips through.
Serious concern detection is configured per-organization. Click the settings icon in the tool header to go directly to Serious Concern Detection settings.
Flagged feedback
Responses flagged by automated moderation rules or by a team member. Useful for tracking edge cases and content policy issues separately from serious concerns.
Form cloud (forms only)
Available when your organization has forms. Shows a word cloud across all form text responses, scoped to whichever form is selected.
Form analysis (forms only)
Aggregated analytics for structured form responses — charts, distributions, and trends broken down by question and answer option.
AI Assistant
The Brain Chat panel (available from the sidebar) lets you ask questions about your feedback in natural language. Examples:
- "What are the most common complaints this week?"
- "Summarize the negative feedback from last month"
- "Are there any patterns in the serious concerns?"
The assistant reads your filtered feedback data and responds with analysis, themes, and suggested actions.
Exporting data
Use the Export button to download responses as a CSV file. The export respects any active filters — only the responses currently shown are included.