Back to results
US 2026/0278412 A1Published application
Offline inference performance analysis of agent workflows
AssigneeNVIDIA Corporation
InventorsMichael David Demoret, Dhruv Nandakumar
Life of this patent
Estimating the term…
Abstract
An AI agent system processes multiple AI agent workflows created with different frameworks, providing standardized abstractions and formatting functions. The toolkit generates instrumentation data and performance efficiency metrics for the workflows based on execution events, and outputs these metrics in a standardized format, allowing for efficient comparison and analysis of the workflows' performance.
Claims20
One or more processors comprising processing circuitry to perform operations comprising:
- receiving, by an artificial intelligence (AI) agent system, one or more AI agent workflows created using different AI agent frameworks, the AI agent system configured to provide AI framework-agnostic abstractions of the one or more AI agent workflows by wrapping individual functions with standardized schemas and descriptions;
- generating, by the AI agent system, instrumentation data for the one or more AI agent workflows based on one or more execution events of the one or more AI agent workflows;
- and
- adjusting, by the AI agent system, a deployment configuration for the one or more AI agent workflows based on one or more performance efficiency metrics generated from the instrumentation data.
1 more element
A system comprising one or more processors to perform operations comprising:
- receiving, by an artificial intelligence (AI) agent system, one or more AI agent workflows created using different AI agent frameworks, the AI agent system configured to provide AI framework-agnostic abstractions of the one or more AI agent workflows by wrapping individual functions with standardized schemas and descriptions;
- generating, by the AI agent system, instrumentation data for the one or more AI agent workflows based on one or more execution events of the one or more AI agent workflows;
- and
- adjusting, by the AI agent system, a deployment configuration for the one or more AI agent workflows based on one or more performance efficiency metrics generated from the instrumentation data.
1 more element
A method comprising:
- receiving, by an artificial intelligence (AI) agent system, one or more AI agent workflows created using different AI agent frameworks, the AI agent system configured to provide AI framework-agnostic abstractions of the one or more AI agent workflows by wrapping individual functions with standardized schemas and descriptions;
- generating, by the AI agent system, instrumentation data for the one or more AI agent workflows based on one or more execution events of the one or more AI agent workflows;
- and
- modifying, by the AI agent system, the one or more AI agent workflows based on one or more performance efficiency metrics generated from the instrumentation data.
1 more element
Dependencies are inferred from the claim text.