Tenant Workload Sequencing for Predictive Service Provisioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computing systems face challenges in determining the appropriate services to provide to tenants and managing their usage effectively, as existing methods are costly, inaccurate, and fail to predict future adoption or growth potential.
Innovation Solution
A control system that generates a model based on attributes of high-performing tenants to predict the likelihood of adoption and usage of workloads or features, allowing for targeted service provisioning and resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual surveys are conducted to obtain tenant adoption information, then current tenant status can be obtained, but the method is costly and inaccurate
Solution Approach 1:
The system automatically collects and analyzes tenant usage data from workload logs and telemetry without requiring manual surveys. The tenant usage model autonomously processes available data to generate adoption metrics, eliminating the need for costly manual information gathering while maintaining high accuracy through continuous automated monitoring.
Solution Approach 2:
The patent replaces manual survey methods with an automated computational model that uses machine learning algorithms to analyze tenant usage patterns. This substitution transforms the mechanical process of conducting surveys into an automated data processing system that continuously monitors and predicts tenant adoption based on observed usage behavior.
2Adaptability or versatility
If services are provided to all tenants uniformly, then service coverage is maximized, but resource allocation efficiency decreases
Solution Approach 1:
The system applies different service levels and support strategies to different tenants based on their specific usage patterns and adoption likelihood scores. Instead of uniform service provision, the tenant usage model identifies high-potential tenants who receive targeted engagement resources, while low-potential tenants receive automated self-service options, optimizing resource allocation according to local tenant characteristics.
Solution Approach 2:
The patent dynamically adjusts service parameters such as support priority, engagement frequency, and resource allocation based on the tenant usage model's predictions. Tenants with high adoption likelihood receive enhanced service parameters, while those with low likelihood receive reduced service parameters, allowing the system to maintain adaptability across all tenants while improving overall resource allocation efficiency.
3Reliability
If tenant adoption metrics are updated frequently, then information remains current, but data processing overhead increases
Solution Approach 1:
The system implements periodic batch processing of tenant usage data at scheduled intervals rather than continuous real-time processing. The tenant usage model is retrained and metrics are updated at defined frequencies based on data availability and business needs, reducing processing overhead while maintaining sufficiently current information for effective decision-making.
Solution Approach 2:
The patent processes only the necessary subset of data required for model updates rather than analyzing all available data continuously. The system selectively processes relevant usage events and filters out redundant information, performing partial processing that maintains metric reliability while minimizing unnecessary computational overhead.
Data Source
AI summary
A control system controls tenant services to various tenants by obtaining tenant attributes for each tenant, with respect to a particular workload. A model is generated that models tenant usage performance for a set of best performing tenants. The model is then applied to a remainder of the tenants to obtain a metric indicative of a likely tenant capacity for incremental usage of the workload. The control system controls the services provided to the tenant based upon the likelihood of adoption metric.