Federated AI Model Chain Routing for Computational Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional software services that utilize machine learning functionality are unable to efficiently correlate user requests with appropriate machine learning model complexities, leading to compromised output quality or unnecessary resource consumption.
Innovation Solution
A federated AI system that uses a model chain to process requests, where a scoring machine learning model determines the lowest computational complexity model that meets a performance threshold, allowing lower complexity requests to be handled by lower complexity models and reserving resources for higher complexity requests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a single high-complexity machine learning model is used to handle all user requests, then output quality is maintained, but computational resource consumption increases unnecessarily
Solution Approach 1:
The patent segments the machine learning model set into multiple complexity levels (first, second, third models with increasing computational complexity). A scoring model evaluates each request and routes it to the appropriate complexity level, ensuring high-complexity models are only used when necessary while maintaining output quality through selective model deployment.
2Productivity
If a single low-complexity machine learning model is used to reduce resource consumption, then computational efficiency improves, but output quality deteriorates
Solution Approach 1:
The patent implements a dynamic model selection mechanism where the scoring model evaluates each user request and dynamically selects the appropriate machine learning model complexity level. This allows the system to adaptively use low-complexity models for simple requests (improving efficiency) while switching to high-complexity models when request characteristics indicate higher quality requirements (maintaining reliability).
3Use of energy by moving object
If multiple machine learning models of varying complexity are deployed, then resource efficiency improves through selective model usage, but system complexity increases
Solution Approach 1:
The patent introduces a scoring model as an intermediary component that manages the complexity of selecting among multiple machine learning models. The scoring model evaluates user requests and determines the appropriate model complexity level, thereby simplifying the overall system architecture by centralizing the decision-making process rather than requiring complex routing logic throughout the system.
4Reliability
If high-complexity machine learning models are used for all requests, then output quality is ensured, but processing speed decreases due to unnecessary computational overhead
Solution Approach 1:
The patent applies local quality by matching model complexity to request characteristics. The scoring model analyzes each user request and assigns it to a machine learning model with appropriate complexity level - using low-complexity models for simple requests (faster processing) and high-complexity models only when request features indicate quality requirements (ensuring output quality). This localized optimization improves overall processing speed while maintaining quality where needed.
Data Source
AI summary
A federated artificial intelligence system executes machine learning models of a model chain in order of increasing computational complexity to determine a lowest computational complexity model to use to serve a quality response to a user request. A first machine learning model of the model chain performs an inference operation to produce first output based on the user request. A scoring machine learning model determines that the first output fails to meet a threshold. Based on such determination, a second machine learning model of the model chain performs a second inference operation to produce second output based on the user request, in which the second machine learning model has a higher computational complexity than the first machine learning model. The scoring machine learning model determines that the second output meets the threshold, and, based on such determination, the second output is transmitted in response to the user request.