Cloud Decision Platform Automating Metric Feedback Loops

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The development of analytic decision applications is time and resource intensive, requiring manual propagation of operational and business metrics for improvement, and existing solutions lack a unified interface for integrating decisioning technologies and managing their lifecycle effectively.

Innovation Solution

A cloud-based decision management platform that provides a unified interface for developing, configuring, deploying, and managing decision services, allowing for the use of various languages and technologies, with features like automated validation, adaptive learning, and lifecycle management, enabling cost-effective and scalable real-time decision management.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If analytic decision applications are developed using conventional methods with manual processes, then developers can maintain control over the development process, but the development time and resource requirements increase significantly

Engineering Contradiction:
Improvedevelopment efficiencyVSAvoiddevelopment time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The system enables self-service through automated propagation of operational and business metrics back to analytic scientists, eliminating manual intervention. The platform automatically captures metrics, processes them, and feeds them back for model improvement, allowing the system to serve itself without continuous human involvement in the feedback loop.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system implements automated feedback mechanisms where operational and business metrics are automatically captured, processed, and propagated back to analytic scientists. This closed-loop feedback system continuously improves analytic models by delivering relevant performance data without manual intervention, significantly reducing development time while maintaining quality.

Inventive Principle:
Principle #23Feedback

2Adaptability or versatility

If multiple languages and technologies are used for decision services, then versatility and adaptability improve, but integration complexity and deployment difficulty increase

Engineering Contradiction:
Improvetechnology compatibilityVSAvoidintegration complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The platform provides universal support for multiple programming languages and technologies through a unified cloud-based interface. It can ingest, manage, and deploy decision services written in different languages (e.g., Java, C#, Python, R) without requiring separate integration processes, making the system universally compatible while simplifying the user experience.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The cloud-based platform acts as an intermediary layer between diverse decision service technologies and the enterprise IT infrastructure. It provides standardized interfaces and automated integration capabilities that mediate between different programming languages and the underlying systems, reducing integration complexity while maintaining support for multiple technologies.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If manual propagation of metrics is used for model improvement, then data accuracy can be maintained through human review, but the time and resources required for model iteration increase

Engineering Contradiction:
Improvedata qualityVSAvoidmodel iteration speed
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The system automatically captures operational and business metrics, processes them through appropriate filters and transformations, and propagates them back to analytic scientists without manual intervention. This self-service approach maintains data quality through automated validation while significantly accelerating the model iteration process.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The platform implements automated feedback loops that continuously capture performance metrics, validate their quality, and deliver them back to model developers. This automated feedback mechanism maintains data reliability through systematic validation while enabling rapid model iteration by eliminating manual data collection and processing steps.

Inventive Principle:
Principle #23Feedback

4Reliability

If comprehensive lifecycle management is implemented for decision services, then system reliability and monitoring capability improve, but the complexity of the management system increases

Engineering Contradiction:
Improvesystem stabilityVSAvoidmanagement system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The cloud-based platform provides comprehensive lifecycle management capabilities including deployment, monitoring, and maintenance of decision services through a single unified interface. It handles multiple management functions (service registration, version control, performance monitoring, automated updates) simultaneously, improving system reliability while presenting a simplified unified interface to users.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3982256B1Cloud-based decision management platform
Publication Date: 2023.11.01 FAIR ISAAC & CO INC
  • EP3982256B1 patent drawingFigure 1
  • EP3982256B1 patent drawingFigure 2
  • EP3982256B1 patent drawingFigure 3

AI summary

A cloud-based decision management platform along with corresponding method, system, and a computer program product are disclosed. At least one component of at least one computing system is selected from a plurality of components of the computing system. The selected component is configured for execution during a runtime of the computing system. The configured component is executed during runtime. The components of the computing system are stored in a catalog module based on at least one characteristic that includes at least one of the following: analytics, decisioning, identity and access management, and optimization.