NoiseScore
An early-warning radar for every floor
Watches every incoming message, call transcript, and review and scores how noisy or unhappy each floor is right now — so ops can step in before the complaint becomes a refund.
- 0–100 Per-floor risk score
- <5 min From signal to alert
- 4 Input channels fused
NoiseScore listens to the concierge channels — WhatsApp threads, call transcripts, in-room tablet requests, and fresh reviews — and extracts signals about disturbance and dissatisfaction. The output is a simple per-floor 0–100 score, updated in minutes.
When a score crosses a threshold, the ops team gets a live nudge with the contributing messages. That gives the duty manager a shot at fixing the problem in the room instead of reading about it in a one-star review a day later.
Everything NoiseScore handles for you
-
Multi-channel ingestion
Reads chat, voice transcripts, tablet tickets and reviews in one pipeline.
-
Per-floor rollup
Scores are rolled up by floor and zone, not just rooms, so clusters surface quickly.
-
Threshold alerts
Breaches push a concise alert to ops with the top contributing messages attached.
-
Drill-down timeline
Tap a floor to see the hour-by-hour signal and the raw source messages behind it.
How NoiseScore plugs into the hotel stack
NoiseScore is a consumer of conversational and review data and a producer of risk events. It does not change state on its own — it nudges other blocks.
- ChatRecorder supplies the full chat and voice transcripts the scorer reads.
- ReviewLoop feeds fresh public reviews into the score with short TTL.
- HousekeepingQueue receives routing hints to avoid cleaning next to a flagged floor.
- AIM CRM shows the current score on the duty-manager dashboard and timeline.
Wire NoiseScore into your product today
Book a consultation with our founders and we'll walk you through the whole microservice stack — not just this one — live on your domain.