System
Value-Based Pricing.
Fortune 500 hospitality operator.
Per-customer pricing at the moment the booking flow asks.
20,000 personalized rates per second. Live.
Constraint
A traditional RMS optimizes the room.
It does not see the customer.
Per-customer, per-property, per-room, per-day.
18.25 trillion rates per pricing run.
Uncomputable in batch.
Compute at request time, or not at all.
Rumi makes it fit.
Foundation
State and execution co-located.
Data is local. Always.
1.8 million rack rates in memory.
10 million customer records in memory.
Pricing logic runs against both in the same node.
Persistence, messaging, and recovery are platform concerns.
Developers write plumbing-free business logic.
Personalization happens inside the request budget.
Outcomes
Pricing budget: 1 ms.
Customer context included.
1.8 million rack rates, 10 million customers co-located in node.
20,000 personalized rates per second.
Live.
OTA, Loyalty, Property, Contact Center, Front Desk.
One architecture across every channel.
The same architecture extended to the in-house CRS.