Foresight
Cancellation Prediction System
Generated from HotelBooking-PRD.pdf

See which bookings will cancel — before the revenue window closes.

A revenue-manager workspace that scores every new booking for cancellation risk in real time, explains the top drivers, and recommends the intervention that protects revenue. Built on a Random Forest model with 87.9% test accuracy.

87.9%
model test accuracy
< 2s
scoring latency
33→25%
cancellation-rate target
$174B
online booking market '26

Explore the screens

Eleven connected views. Every screen links to the others — click in and navigate like the real product.

How it works

The five PRD user flows, end to end.

Who it's for

Revenue Manager

Primary user. Works the at-risk queue, logs interventions, tunes the risk threshold.

General Manager

Weekly reporting — cancellation forecast, intervention ROI, model accuracy, segment policy.

Front Desk / Ops

Receives advisory overbooking buffers and high-risk arrival notifications for check-in prep.

PRD traceability

Each screen maps back to specific requirements in the PRD.

ScreenServesPRD reference