Decoupling Decision Engines from Applications for Rapid Model Deployment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The tightly coupled nature of application programs with their data and decision-making functionality hampers agility in deploying updates and changes, especially in large-scale applications serving thousands of users, leading to inefficiencies in deployment cycles and reduced flexibility.
Innovation Solution
A system and method that decouples decision engines from application programs, allowing for the deployment of predictive models using electronic data from multiple sources, enabling rapid and flexible deployment of consumer-specific decisions with reduced processor time and scalable solutions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If application programs maintain tight coupling with decision-making functionality, then reliability and integration are improved, but agility and deployment speed deteriorate
Solution Approach 1:
The patent segments the monolithic application program into separate components: the application program itself and the predictive model library. This segmentation allows the predictive models to be independently developed, tested, and deployed without affecting the application program, thereby improving deployment speed while maintaining integration reliability through defined interfaces.
Solution Approach 2:
The patent extracts the decision-making functionality (predictive models) from the application program and places it in a separate, externally accessible library. This extraction enables the models to be updated and deployed independently, resolving the contradiction between maintaining tight integration for reliability and enabling rapid deployment for productivity.
2Adaptability or versatility
If application programs are updated frequently to improve functionality, then adaptability is improved, but system stability and user impact worsen
Solution Approach 1:
By segmenting the system into application programs and separate predictive model libraries, the patent enables frequent updates to model functionality without requiring application program updates. This maintains system stability while improving functional adaptability, as models can be swapped independently.
Solution Approach 2:
The patent introduces dynamic model selection and switching capabilities, allowing the system to adapt functionality by loading different predictive models based on needs without restructuring the application program. This dynamic approach enables adaptability while preserving system stability through controlled model deployment.
3Measurement precision
If decision models require significant processing resources, then measurement precision is improved, but resource efficiency and scalability worsen
Solution Approach 1:
The patent implements partial loading of predictive models, where only the necessary models are loaded into memory based on specific application needs. This allows high-precision models to be maintained with full accuracy when needed, while reducing resource consumption by not loading all models simultaneously, thus balancing measurement precision with resource efficiency.
Solution Approach 2:
The patent introduces a temporal dimension to model loading, allowing models to be loaded and unloaded dynamically based on demand. This enables the system to achieve high decision accuracy when models are active while minimizing resource usage during idle periods, effectively resolving the contradiction between precision and resource efficiency.
Data Source
AI summary
A system and method for deploying predictive models is presented which personalizes application program functionality and user interface presentation to the consumer user. That personalization takes place within a decision engine that decouples decision models from application processes so that rapid deployment may take place. When a consumer user specific decision is required, the application program interacts with the decision model to trigger a consumer-user appropriate decision tailored to characteristics of the consumer user. New models are introduced into a decision process without requiring any application program change.