US 12,730,927 B2Granted

Adaptable framework for integration of generative artificial intelligence

AssigneeAmerican Express Travel Related Services Company, Inc.

InventorsPedro Burglin Paes, Andras L. Ferenczi, Hiranmayi Palanki, Mukund Shankar Simharaghu

Life of this patent

Estimating the term…

Loading patent…

Abstract

Disclosed herein are system, method, and computer program product embodiments for managing generative artificial intelligence (AI) queries from client systems and responses to the queries from generative AI systems. A request management system may provide an adaptable framework for receiving, managing, monitoring, and/or controlling generative AI queries received from client systems seeking generative AI content. The request management system may authenticate client systems and generative AI systems. The request management system may also administer security and observabilities protocols to the queries and responses. The request management system may identify a ruleset that includes one or more conditions that indicate whether the queries and responses may be forwarded to the generative AI systems and client systems, respectively. This may provide enterprise computing control over interactions between client devices and generative AI systems.

Claims20

Dependent claims

A computer implemented method, comprising:

receiving, from a client system, a generative artificial intelligence (AI) query that requests a response from one or more generative AI systems;
conducting, using a first API gateway, a first authentication of the client system;
in response to the first authentication, evaluating, by a request management system content and HTTP metadata of the generative AI query, wherein the request management system comprises a synchronous control and an asynchronous control, wherein the synchronous control is configured to determine when the content of the generative AI query satisfies one or more conditions of a ruleset and the asynchronous control is configured to asynchronously update a machine learning model based on the content of the generative AI query, wherein the machine learning model is configured to determine the one or more conditions of the ruleset;
identifying, using the synchronous control, the one or more conditions of the ruleset for forwarding generative AI queries to the one or more generative AI systems;
determining, based on the identifying, the content of the generative AI query satisfies the one or more conditions of the ruleset;
updating, using the asynchronous control, the machine learning model that determines the one or more conditions of the ruleset;
in response to the determining, conducting a second authentication, using a second API gateway and a key storage system, of the one or more generative AI systems;
in response to the second authentication, transmitting the generative AI query from the request management system to the one or more generative AI systems;
and
evaluating, by the one or more generative AI systems, the content of the generative AI query from the request management system.

7 more elements

A system, comprising:

a memory;
and
at least one processor coupled to the memory and configured to:
receive, from a client system, a generative artificial intelligence (AI) query that requests a response from one or more generative AI systems;
conduct, using a first API gateway, a first authentication of the client system;
in response to the first authentication, evaluating, by a request management system, content and HTTP metadata of the generative AI query, wherein the request management system comprises a synchronous control and an asynchronous control, wherein the synchronous control is configured to determine when the content of the generative AI query satisfies one or more conditions of a ruleset and the asynchronous control is configured to asynchronously update a machine learning model based on the content of the generative AI query, wherein the machine learning model is configured to determine the one or more conditions of the ruleset;
identify, using the synchronous control, the one or more conditions of the ruleset for forwarding generative AI queries to the one or more generative AI systems;
determine, based on the identification, the content of the generative AI query satisfies the one or more conditions of the ruleset;
update, using the asynchronous control, the machine learning model that determines the one or more conditions of the ruleset;
in response to the determination, conduct a second authentication, using a second API gateway and a key storage system, of the one or more generative AI systems;
in response to the second authentication, transmit the generative AI query from the request management system to the one or more generative AI systems;
and
evaluate, by the one or more generative AI systems, the content of the generative AI query from the request management system.

10 more elements

A non-transitory computer-readable device having instructions stored thereon that, when executed by at least one computing device, cause the at least one computing device to perform operations comprising:

receiving, from a client system, a generative artificial intelligence (AI) query that requests a response from one or more generative AI systems;
conducting, using a first API gateway, a first authentication of the client system;
in response to the first authentication, evaluating, by a request management system, content and HTTP metadata of the generative AI query, wherein the request management system comprises a synchronous control and an asynchronous control, wherein the synchronous control is configured to determine when the content of the generative AI query satisfies one or more conditions of a ruleset and the asynchronous control is configured to asynchronously update a machine learning model based on the content of the generative AI query, wherein the machine learning model is configured to determine the one or more conditions of the ruleset;
identifying, using the synchronous control, the one or more conditions of the ruleset for forwarding generative AI queries to the one or more generative AI systems;
determining, based on the identifying, the content of the generative AI query satisfies the one or more conditions of the ruleset;
updating, using the asynchronous control, the machine learning model that determines the one or more conditions of the ruleset;
in response to the determining, conducting a second authentication, using a second API gateway and a key storage system, of the one or more generative AI systems;
in response to the second authentication, transmitting the generative AI query from the request management system to the one or more generative AI systems;
and
evaluating, by the one or more generative AI systems, the content of the generative AI query from the request management system.

7 more elements

Dependencies are inferred from the claim text.

Facts

Application: 18/391,044Filed 20 Dec 2023
Published: US 2025/0209200 A126 Jun 2025
Granted: 8 Sep 2026
Term ends: Not estimatedThe source records don’t support an estimate.

Classification8 codes

G06F 21/6245
To a system of files or objects, e.g. local or distributed file system or database: protecting personal data, e.g. for financial or medical purposes
G06F 21/1064
Enforcing content protection by specific content processing: restricting content processing at operating system level
G06F 21/44
Authentication, i.e. establishing the identity or authorisation of security principals: program or device authentication
G06N 20/00
Machine learning