AI Insights
Automatically generated summaries and analysis of your feedback data, powered by Google Gemini.
Last updated on
AI Insights is Enafeedback's automated analysis feature. It reads your recent feedback and generates concise, actionable summaries using Google Gemini.
What AI Insights analyses
Insights can be generated for each feedback module:
| Module | What the insight covers |
|---|---|
| Hygiene | Dominant low-score categories, repeat problem locations, trend direction |
| Survey | Key themes in text responses, outlier scores, NPS movement explanation |
| Feedback (Tickets) | Most common complaint categories, unresolved high-priority issues |
| Cleaning | Compliance gaps, frequently missed tasks, overdue locations |
Generating an insight
- Navigate to any module (e.g., Hygiene).
- Apply date filters to select the period you want analysed.
- Click Generate AI Insight (sparkle icon, top right).
- Wait 5–15 seconds while the insight is generated.
- The insight appears as a collapsible card at the top of the page.
Insights are cached — generating an insight for the same module and date range twice shows the same result without consuming additional AI quota. The cache expires after 24 hours.
Insight content
A typical insight contains:
- Summary — 2–3 sentence overview of the period.
- Key findings — Bullet points of the most significant patterns.
- Problem areas — Specific locations or categories that underperformed.
- Positive highlights — Areas performing well.
- Suggested action — One or two specific recommendations.
Privacy and PII handling
Before any feedback text is sent to the Google Gemini API:
- All personally identifiable information (names, emails, phone numbers, addresses) is detected and replaced with
[REDACTED]tokens using a server-side anonymisation step. - The anonymised text is then sent to Gemini.
- The returned insight is stored in the workspace database — it never leaves your region.
This process is automatic and requires no configuration.
AI Insights comply with GDPR and KVKK requirements. No raw personal data is transmitted to external AI services.
Insight history
All generated insights are stored and accessible in the AI Insights history panel. You can:
- View past insights by module and date range
- Compare insights across periods
- Export an insight as PDF (useful for weekly stakeholder reports)
Rate limiting
Insight generation is subject to your plan's AI quota (measured in tokens). Each insight generation consumes approximately 2,000–8,000 tokens depending on the volume of feedback data. See AI Usage & Quotas for details.