US 12,476,979 B2
Agentically-orchestrated foundational models for cloud security controls
Aayush Srivastava, New York, NY (US); Alison Marlene Andrews, Stonington, CT (US); and Sai Rama Raju Bhupathiraju, Cedar Park, TX (US)
Assigned to GOOGLE LLC, Mountain View, CA (US)
Filed by Google LLC, Mountain View, CA (US)
Filed on May 1, 2024, as Appl. No. 18/651,887.
Prior Publication US 2025/0343799 A1, Nov. 6, 2025
Int. Cl. H04L 9/40 (2022.01)
CPC H04L 63/105 (2013.01) 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
obtaining, by a computing system comprising one or more computing devices, cloud architecture information descriptive of a proposed cloud architecture, comprising:
user response information comprising information indicative of cloud architecture requirements for the proposed cloud architecture to fulfill; and
component selection information indicative of a plurality of cloud components that meet cloud architecture requirements for the proposed cloud architecture;
processing, by the computing system, the cloud architecture information with a plurality of agentic orchestration models to obtain a role output comprising security control information indicative of a plurality of security controls selected for the proposed cloud architecture, wherein each of the plurality of agentic orchestration models comprises a machine-learned model prompted to fulfill a corresponding cloud architecting role of a plurality of cloud architecting roles, and wherein the role output is associated with a cloud security role of the plurality of cloud architecting roles; and
based on the security control information, providing, by the computing system, information indicative of the plurality of security controls selected for the proposed cloud architecture to a user device associated with the user response information.