Actuals Cache for Revenue Management Analytics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional revenue management systems in the travel and hospitality industry fail to effectively utilize user travel data for actionable insights and analysis, leading to suboptimal decision-making and profit opportunities.
Innovation Solution
An actuals cache system for a revenue management system analytics engine that utilizes a pipeline of data services, including two-dimensional dates and caching infrastructure, to process and analyze large amounts of data efficiently, providing insights such as demand forecasts and pricing recommendations by leveraging a NoSQL database and cache management modules.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional revenue management systems process user travel data in real-time, then analysis accuracy is improved, but system performance and response time deteriorate
Solution Approach 1:
The patent implements an ETL (Extract, Transform, Load) pipeline that pre-processes user travel data during off-peak hours before analytics queries are executed. This preliminary action transforms raw data into structured, query-optimized formats and pre-calculates aggregate metrics, enabling fast retrieval during runtime without compromising analysis accuracy
Solution Approach 2:
The system applies different processing qualities to different data elements: frequently accessed aggregate metrics are pre-computed and cached with high optimization, while detailed individual records maintain their original structure for accuracy. This local differentiation allows the system to optimize performance for common queries without losing analytical precision
2Loss of information
If conventional revenue management systems store all raw travel data, then data completeness is improved, but data retrieval and processing time worsen
Solution Approach 1:
The patent segments the data storage system into multiple layers: a compressed raw data warehouse for complete data retention, a structured intermediate layer for transformed data with optimized schemas, and a cache layer for frequently accessed aggregates. This segmentation allows the system to retrieve only the necessary data portion for each query type, dramatically reducing retrieval time while maintaining access to complete raw data when needed
Solution Approach 2:
The ETL pipeline acts as an intermediary between the raw data warehouse and analytics engines. It transforms and pre-aggregates data into intermediate structures that are optimized for specific query patterns, serving as a mediator that reduces the complexity of querying raw data while preserving the ability to access complete information through the pipeline
3Adaptability or versatility
If revenue management systems execute complex analytics queries on raw data, then analysis depth is improved, but processing speed deteriorates
Solution Approach 1:
The system pre-computes and caches aggregate metrics, rolling totals, and common analytical views during the ETL phase. When analytics queries are executed, these pre-computed results are retrieved and combined with current data, enabling deep analysis without re-processing raw data for every query
Solution Approach 2:
The patent introduces a temporal dimension to data processing by maintaining historical snapshots and versioned data states. This allows analytics to operate on pre-processed historical data while current data is being updated, effectively adding a time dimension that separates processing speed from analysis depth
Data Source
AI summary
At a data ingestion time, a travel sessionizer captures raw event data representing a plurality of bookings made by a plurality of users for a given day, identifies a plurality of cache entries corresponding to a travel property for the given day, wherein at least one of the plurality of bookings from the raw event data is for the travel property, and updates the plurality of cache entries corresponding to the travel property for the given data to reflect the at least one booking from the raw event data. At a runtime, the travel sessionizer receives a travel analytic query corresponding to the travel property, accesses at least one of the plurality of cache entries corresponding to the travel property for the given day, and executes the travel analytic query using travel data from the at least one of the plurality of cache entries.