IP Library › Granted Patent US 12,260,260
Granted Patent B1
US 12,260,260 · App. 18/622,721 · Granted Mar 25, 2025

Digital delegate computer system architecture for improved multi-agent large language model (LLM) implementations

Inventors: Matthew J. Gorman (Burlington, CT); Vincent E. Haines (East Point, GA); Girish A. Modgil (Alpharetta, GA); Brad E. Gawron (Porter Corners, NY)
Assignee: The Travelers Indemnity Company
G06F9/5066G06F9/468G06F9/4881G06F2209/485G06F2209/5017G06F2209/506
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 12,260,260
App. No.
18/622,721
Granted
Mar 25, 2025
Kind
B1
Abstract

Systems, apparatus, methods, and articles of manufacture for digital delegate computer system architecture that provides for improved multi-agent LLM implementations.

Claims (30)

1. A multi-Large Language Model (LLM), multi-agent, digital delegate computer system, comprising:

an identity server comprising at least one electronic processing device;

a non-transitory access entitlement data store device in communication with the identity server, the access entitlement data store device storing access entitlement data in relation to user identification data;

a multi-agent LLM server comprising a plurality of electronic processing devices and being in communication with the identity server; and

a non-transitory LLM data store device in communication with the multi-agent LLM server, the LLM data store device storing (i) instructions defining a primary LLM agent, (ii) instructions defining a primary LLM; (iii) data descriptive of a plurality of LLM tools, (iv) instructions defining, for each LLM tool of the plurality of LLM tools, a secondary LLM agent, and (v) operating instructions, that when executed by the plurality of electronic processing devices, result in:

receiving, by the multi-agent LLM server and from a user device, a user request comprising a prompt and an indication of an identifier of a user of the user device;

identifying, by the identity server and utilizing the identifier of the user to query the access entitlement data store, at least one access entitlement assigned to the user;

identifying, by an execution of the primary LLM agent by the multi-agent LLM server, and based on the at least one access entitlement assigned to the user, and by querying the data descriptive of the plurality of LLM tools, a subset of LLM tools from the plurality of LLM tools that the user is entitled access to;

generating, by the primary LLM and utilizing both the prompt and the identified subset of LLM tools from the plurality of LLM tools that the user is entitled access to, and after the identification of the subset of LLM tools from the plurality of LLM tools that the user is entitled access to, a multi-tier plan for responding to the user request, wherein the multi-tier plan defines a plurality of actions, with each action being assigned to one of the LLM tools from the identified subset of LLM tools;

executing, by the primary LLM agent, the multi-tier plan, by:

(i) calling a secondary LLM agent assigned to each respective one of the LLM tools from the identified subset of tools for the plurality of actions of the multi-tier plan; and

(ii) receiving, from each secondary LLM agent and in response to the calling, a response for each of the actions of the multi-tier plan;

constructing, by the primary LLM agent and utilizing the responses for the actions of the multi-tier plan, a user response; and

transmitting, by the primary LLM agent and to the user device, the user response.

2. The multi-LLM, multi-agent, digital delegate computer system of claim 1 , wherein the execution by the plurality of electronic processing devices further results in:

transmitting, by the multi-agent LLM server and to the identity server, the identifier of the user; and

wherein the identifying of the at least one access entitlement assigned to the user is conducted in response to the transmitting of the identifier of the user by the multi-agent LLM server.

3. The multi-LLM, multi-agent, digital delegate computer system of claim 1 , wherein the at least one access entitlement assigned to the user comprises a suite of access entitlements defined by a role assigned to the user.

4. The multi-LLM, multi-agent, digital delegate computer system of claim 1 , wherein each LLM tool from the subset of tools from the plurality of LLM tools that the user is entitled access to comprises a different secondary LLM.

5. The multi-LLM, multi-agent, digital delegate computer system of claim 4 , wherein the execution by the plurality of electronic processing devices further results in:

executing, in response to the calling and by each of the called secondary LLM agents, a respective secondary LLM.

6. The multi-LLM, multi-agent, digital delegate computer system of claim 1 , wherein the multi-tier plan is generated as a text file.

7. The multi-LLM, multi-agent, digital delegate computer system of claim 1 , wherein the generating of the multi-tier plan is based on a subset of the data descriptive of the plurality of LLM tools that corresponds to the identified subset of LLM tools from the plurality of LLM tools that the user is entitled access to.

8. The multi-LLM, multi-agent, digital delegate computer system of claim 7 , wherein the subset of the data that corresponds to the identified subset of LLM tools from the plurality of LLM tools that the user is entitled access to comprises at least one of: (i) a cost of each LLM tool, (ii) a bandwidth of each LLM tool, (iii) a rating of each LLM tool, and (iv) a historic performance metric of each LLM tool.

9. The multi-LLM, multi-agent, digital delegate computer system of claim 1 , wherein each action of the plurality of actions of the multi-tier plan is defined by the primary LLM based on a different goal derived by the primary LLM from the prompt.

10. The multi-LLM, multi-agent, digital delegate computer system of claim 9 , wherein each LLM tool assigned to each action of the plurality of actions of the multi-tier plan is selected by the primary LLM based on a stored indication of an ability of each LLM tool to handle the respective assigned action.

11. The multi-LLM, multi-agent, digital delegate computer system of claim 10 , wherein the stored indication of the ability of each LLM tool to handle the respective assigned action is derived from previous performance data for the respective LLM tool.

12. The multi-LLM, multi-agent, digital delegate computer system of claim 10 , wherein the stored indication of the ability of each LLM tool to handle the respective assigned action comprises at least one of a score and a ranking.

13. The multi-LLM, multi-agent, digital delegate computer system of claim 1 , wherein the user response is utilized to update at least a portion of the data descriptive of a plurality of LLM tools stored in the non-transitory LLM data store device.

14. The multi-LLM, multi-agent, digital delegate computer system of claim 1 , wherein the responses for the actions of the multi-tier plan are utilized to update at least a portion of the data descriptive of a plurality of LLM tools stored in the non-transitory LLM data store device.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 30, 2024
From: GORMAN, MATTHEW J.; HAINES, VINCENT E.; MODGIL, GIRISH A.; GAWRON, BRAD E.
To: THE TRAVELERS INDEMNITY COMPANY
Reel/Frame 068122/0023 →
References Cited (8)
US 11640823B1 · Pemberton · 2023 [cited by examiner]
US 11928426B1 · Gutzeit · 2024 [cited by examiner]
US 20240146734A1 · Southgate · 2024 [cited by examiner]
US 20240202225A1 · Siebel · 2024 [cited by examiner]
US 20240202460A1 · Schillace · 2024 [cited by examiner]
US 20240346388A1 · Wilczynski · 2024 [cited by examiner]
Ding, Tinghe. “MobileAgent: enhancing mobile control via human-machine interaction and SOP integration.” arXiv preprint arXiv: 2401.04124 (2024). (Year: 2024). [cited by examiner]
Website: https://epam-rail.com/platform; download date Jun. 6, 2024; 2 pps. [cited by applicant]
Cited By (3)
US 12,518,109 US 12,547,677 US 12,719,888