Catering Demand Forecasting demo
Forecasts meals per flight from bookings, cabin mix and past use, for catering planners 1:33
What the demo shows
Forecasts meals per flight from bookings, cabin mix and past use, for catering planners
Agents forecast how many passengers will board each cabin from bookings, no-show history and upgrades, remove connecting passengers already rebooked off late inbound flights, and forecast the dish mix from pre-orders and route history. Before each caterer cut-off the Order Planner proposes the final order in one sentence with its reasons and sources, a second agent re-derives the forecast independently, and the catering planner approves, changes or rejects it. Only approved changes go to the caterer, whose acknowledgment is tracked. After each flight the agents read the return-load report, measure waste and run-outs against their own forecast and show their accuracy openly.
Transcript
- Catering Demand Forecasting tells catering planners how many meals each flight needs.
- Caterers load bookings plus a buffer, and untouched food is a quarter of cabin waste.
- Rene opens Corvain 118 to Frankfurt.
- 41 business passengers booked.
- Corvain 902 from Boston is late.
- For connecting, passengers will miss this flight.
- Every number opens its source.
- Here, the flight's staff.
- status. 37 will board.
- 61% will choose the short rib.
- The proposal. 40 meals, not 46, saving $612.
- A second agent checks the forecast from raw data and agrees.
- Rene approves 40 with a reason.
- The caterer confirms two minutes later.
- After every flight, the agents grade their forecasts in public.
- Every value cited, every change approved, every step logged.
- The result? An estimated $7.5 million a year saved.
- Built on Botminds: agents, data, screens, and governance on one platform.
Transcribed from the film’s narration. Captions are on the player.