IP Library Granted Patent US 12689524
Granted Patent B1
US 12689524 · App. 19/538,785 · Granted Jul 21, 2026

Selecting artificial intelligence agents for function execution

Inventors: Ganesh Prasad Bhat (West Orange, NJ); Imir Arifi (Nashville, TN); James Myers (New York, NY); Sofia Rahman (New York, NY)
Assignee: Citibank, N.A.
H04L9/3247H04L9/50G06F21/62H04L9/3297H04L63/308
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12689524
App. No.
19/538,785
Granted
Jul 21, 2026
Kind
B1
Abstract

Methods and systems for routing execution requests to autonomous artificial intelligence agents using performance-based selection are disclosed herein. An agent routing system may receive a prompt for completing an action using artificial intelligence agents from a plurality of agents, where each agent may be associated with computer-executable operations configured for autonomous execution on software applications. The system may determine a set of agents enabled to complete the requested action and may access action completion datasets for these agents. Each dataset may include action parameters for completed actions, including status parameters indicating successful completion and timestamps of completion. The system may select an artificial intelligence agent based on a frequency composite parameter generated from successful completion status parameters and a time composite parameter indicating completion timing. The system may identify parameters required by the selected agent to complete the action and may transmit these parameters to the artificial intelligence agent for execution.

Claims (63)

1 . A system comprising:

at least one hardware processor; and

at least one non-transitory memory storing instructions, which, when executed by the at least one hardware processor, cause the system to:

obtain an instruction to complete an action using one or more artificial intelligence agents of a plurality of artificial intelligence agents;

access a set of action completion datasets for the one or more artificial intelligence agents,

wherein each action completion dataset comprises a corresponding plurality of action parameters for a plurality of actions completed by each of the one or more artificial intelligence agents;

select, based on the corresponding plurality of action parameters, a particular artificial intelligence agent of the one or more artificial intelligence agents to complete the action;

cause execution of the action by transmitting a corresponding set of execution parameters associated with the action to the particular artificial intelligence agent;

receive a message from the particular artificial intelligence agent from the plurality of artificial intelligence agents comprising a completion status and a cryptographic signature;

verify the message using a public key associated with the particular artificial intelligence agent; and

cause transmission of an operation that indicates the completion status associated with the action to a node, wherein the node is configured to commit the operation to a distributed database.

2 . The system of claim 1 , wherein the instructions further cause the system to:

update at least one action completion dataset based on the completion status associated with the action.

3 . The system of claim 1 , wherein the instructions further cause the system to:

compare the corresponding plurality of action parameters of the one or more artificial intelligence agents to one or more criterion associated with the action.

4 . The system of claim 1 , wherein the distributed database comprises a blockchain.

5 . A non-transitory computer-readable storage medium comprising instructions stored thereon, wherein the instructions when executed by at least one data processor of a system, cause the system to:

access an instruction to complete an action using at least one artificial intelligence agent;

identify an artificial intelligence agent set configured to complete the action;

determine, based on a corresponding action parameter set for a corresponding action set completed by each of the artificial intelligence agent set, a subset of the artificial intelligence agent set to complete the action;

cause transmission of a corresponding set of execution parameters associated with the action to the subset of the artificial intelligence agent set;

obtaining a message from a particular artificial intelligence agent from the subset of the artificial intelligence agent set comprising a completion status and a cryptographic signature;

verify the message using a public key associated with the particular artificial intelligence agent; and

cause transmission of an operation that indicates the completion status associated with the action to a distributed database.

6 . The non-transitory computer-readable storage medium of claim 5 , wherein the completion status comprises one or more of:

an indication of whether the action was completed according to one or more criterion; or

a timestamp that indicates when the action was completed.

7 . The non-transitory computer-readable storage medium of claim 5 , wherein the completion status comprises one or more of:

generate a cryptographically signed record indicating the completion status; and

cause transmission of the cryptographically signed record to a node of the distributed database.

8 . The non-transitory computer-readable storage medium of claim 5 , wherein the instructions further cause the system to:

determine a performance score for each artificial intelligence agent in the artificial intelligence agent set; and

determine the subset of the artificial intelligence agent set based on a corresponding performance score.

9 . The non-transitory computer-readable storage medium of claim 5 , wherein the instructions further cause the system to:

detect an availability of a new artificial intelligence agent different from the artificial intelligence set; and

store, in the distributed database, an identifier associated with the new artificial intelligence agent.

10 . The non-transitory computer-readable storage medium of claim 5 , wherein the instructions further cause the system to:

compare the corresponding action parameter set of the artificial intelligence agent set to one or more criterion associated with the action.

11 . The non-transitory computer-readable storage medium of claim 5 , wherein the distributed database comprises a distributed ledger.

12 . A computer-implemented method comprising:

storing, in a distributed database, historical action completion data associated with a plurality of artificial intelligence agents;

receiving a request to perform an action;

generating a selection of a selected artificial intelligence agent capable of performing the action based on the historical action completion data;

causing transmission of, in response to the selection, execution data associated with the action to the selected artificial intelligence agent;

obtaining a transmittal from the selected artificial intelligence agent indicating a status associated with the performed action;

verifying the transmittal using a key associated with the selected artificial intelligence agent; and

causing transmission of the status associated with the performed action to the distributed database.

13 . The computer-implemented method of claim 12 , wherein the historical action completion data comprises a plurality of records, each record corresponding to a previously completed action.

14 . The computer-implemented method of claim 12 , further comprising:

detecting an availability of a new artificial intelligence agent; and

storing an identifier associated with the new artificial intelligence agent in the distributed database.

15 . The computer-implemented method of claim 12 , further comprising:

generating a cryptographic signature for the status associated with the performed action prior to causing transmission to the distributed database.

16 . The computer-implemented method of claim 12 , further comprising:

comparing the historical action completion data of the plurality of artificial intelligence agents to one or more criterion associated with the action.

17 . The computer-implemented method of claim 13 , further comprising:

updating the historical action completion data based on the status associated with the performed action.

18 . The computer-implemented method of claim 12 , wherein the distributed database comprises a distributed ledger.

19 . The computer-implemented method of claim 12 , further comprising:

determining a score for each artificial intelligence agent in the plurality of artificial intelligence agents; and

determining the selected artificial intelligence agent based on a corresponding score.

20 . The computer-implemented method of claim 12 , wherein the selection of a selected artificial intelligence agent is generated based on a corresponding action parameter set for a corresponding action set completed by each of the plurality of artificial intelligence agents.

21 . The computer-implemented method of claim 20 , further comprising: comparing the corresponding action parameter set of the artificial intelligence agent set to one or more criterion associated with the action.