US-20260127464-A1 - Conversational AI Agentic Architecture for Enterprises

Abstract

An agentic AI system architecture able to respond to a user request includes multiple agents, each agent being assigned to a functional domain required by an enterprise. The agentic AI system architecture also includes at least one reasoning engine supported by at least one large language model (LLM), with the reasoning engine being able to interpret a user request and decompose the user request into one or more mini-tasks. An orchestrator module is connected to the reasoning engine and able to route those mini-tasks based on the user request to at least one of the multiple agents for further processing. Together, the orchestrator module and reasoning engine can collect and reconcile processed results from the multiple agents and prepare a response to the user request.

Inventors

• Antonio Nucci

Assignees

• Aisera, Inc.

Dates

Publication Date

20260507

Application Date

20251031

Claims (16)

1. 1 . An agentic AI system architecture able to respond to a user request, comprising: multiple agents, each agent being assigned to a functional domain required by an enterprise; a reasoning engine supported by at least one large language model (LLM), with the reasoning engine being able to interpret a user request and decompose the user request into one or more mini-tasks; an orchestrator module connected to the reasoning engine and able to route those mini-tasks based on the user request to at least one of the multiple agents for further processing; and wherein the orchestrator module and reasoning engine can collect and reconcile processed results from the multiple agents and prepare a response to the user request.
2. 2 . The agentic AI system architecture of claim 1 , wherein multiple LLMs are used to assist in domain specific interpretation of the user request.
3. 3 . The agentic AI system architecture of claim 1 , wherein the reasoning engine can engage the user for clarification in response to the user request.
4. 4 . The agentic AI system architecture of claim 1 , wherein the reasoning engine supports reinforcement learning.
5. 5 . The agentic AI system architecture of claim 1 , wherein the multiple agents further comprise at least some external agents.
6. 6 . The agentic AI system architecture of claim 1 , wherein functional domains further comprise at least one of IT, HR, Finance, Engineering, and Sales and Marketing.
7. 7 . The agentic AI system architecture of claim 1 , wherein the user request is submitted through an engagement channel to the agentic AI System.
8. 8 . A method of orchestrating actions in an agentic AI system, comprising the steps of: receiving a user request through an engagement channel; using a reasoning engine to analyze the received user request and in a task decomposition step identify one or more mini-tasks that require fulfillment; using an orchestration module connected to the reasoning engine to determine which of multiple agents can be selected to fulfill the mini-tasks; and wherein each agent executes the respective mini-task and sends output to the orchestrator module for result aggregation.
9. 9 . A method of orchestrating actions in the agentic AI system of claim 8 , wherein the orchestration module can act in an unsupervised mode to discover which of multiple agents are needed to fulfill the user request, plan a sequence of agent invocation, execute the plan, and verify execution correctness.
10. 10 . A method of orchestrating actions in the agentic AI system of claim 8 , wherein the orchestration module can act in a supervised mode with mini-tasks and agents used being externally provided.
11. 11 . A method of orchestrating actions in the agentic AI system of claim 8 , wherein the orchestration module can act in a semi-supervised mode with mini-tasks being externally provided and the agents used being determined by the agentic AI system.
12. 12 . A method of orchestrating actions in the agentic AI system of claim 8 , wherein each agent independently executes the respective mini-task.
13. 13 . A method of orchestrating actions in the agentic AI system of claim 8 , wherein agents can interact with other agents to execute the respective mini-task.
14. 14 . A method of orchestrating actions in the agentic AI system of claim 8 , wherein each agent is assigned to a functional domain required by an enterprise.
15. 15 . A method of operating an agentic AI system able to respond to a user request, comprising the steps of: selecting a set of multiple agents, each agent being assigned to a functional domain required by an enterprise; providing a reasoning engine supported by at least one large language model (LLM), with the reasoning engine acting to interpret a user request and decompose the user request into one or more mini-tasks; providing an orchestrator module connected to the reasoning engine and routing mini-tasks based on the user request to at least one of the multiple agents for further processing; and wherein the orchestrator module and reasoning engine can collect and reconcile processed results from the multiple agents and prepare a response to the user request.
16. 16 . A method of orchestrating actions in the agentic AI system of claim 15 , wherein agents are dynamically added or removed from the set of multiple agents.

Description

RELATED APPLICATION The present disclosure is part of a non-provisional patent application claiming the priority benefit of U.S. Patent Application No. 63/715,057, filed on November 1, 2024, which is hereby incorporated by reference in its entirety. TECHNICAL FIELD The present disclosure generally relates to use of a system and process for enabling a conversation capable artificial intelligence based system that utilizes multiple connected agents to support autonomous decision making and actions. BACKGROUND The evolution of AI within enterprises has marked a transition from rigid, rule-based systems to more advanced, flexible architectures capable of understanding and executing complex tasks. Non-agentic AI systems, which rely on fixed workflows and predefined rules and experiences, have proven inadequate in addressing the dynamic needs of modern enterprises. Agentic AI architecture represents a transformative approach in the field of artificial intelligence, enabling the development and deployment of autonomous agents capable of dynamic learning, decision-making, and interaction within complex environments. Enterprises adopting Agentic AI architecture can achieve significant benefits, including enhanced operational efficiency, improved decision-making, and more personalized user interactions. By automating complex tasks and continuously learning from real-time data, Agentic AI systems can optimize workflows, reduce operational costs, and respond more effectively to changing business needs. This adaptability not only drives higher productivity but also enables organizations to stay competitive in an increasingly dynamic marketplace. SUMMARY In some embodiments, an agentic AI system architecture able to respond to a user request includes multiple agents, each agent being assigned to a functional domain required by an enterprise. The agentic AI system architecture also includes at least one reasoning engine supported by at least one large language model (LLM), with the reasoning engine being able to interpret a user request and decompose the user request into one or more mini-tasks. An orchestrator module is connected to the reasoning engine and able to route those mini-tasks based on the user request to at least one of the multiple agents for further processing. Together, the orchestrator module and reasoning engine can collect and reconcile processed results from the multiple agents and prepare a response to the user request. In one embodiment, a method of orchestrating actions in an agentic AI system includes the steps of receiving a user request through an engagement channel. In a next step, a reasoning engine is used to analyze the received user request and in a task decomposition step identify one or more mini-tasks that require fulfillment. An orchestration module connected to the reasoning engine is used to determine which of multiple agents can be selected to fulfill the mini-tasks and each agent executes the respective mini-task and sends output to the orchestrator module for result aggregation. In one embodiment, a method of operating an agentic AI system able to respond to a user request includes the steps selecting a set of multiple agents, each agent being assigned to a functional domain required by an enterprise. A reasoning engine supported by at least one large language model (LLM) is provided, with the reasoning engine acting to interpret a user request and decompose the user request into one or more mini-tasks. An orchestrator module connected to the reasoning engine provides for routing of mini-tasks based on the user request to at least one of the multiple agents for further processing. The orchestrator module and reasoning engine can collect and reconcile processed results from the multiple agents and prepare a response to the user request. In some embodiments, the agents can be dynamically added or removed from the set of multiple agents. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 depicts an agentic AI system architecture able to respond to user requests in accordance with an embodiment. FIG. 2 depicts a method of operating an agentic AI system in accordance with an embodiment. FIG. 3 depicts a computer system capable of supporting or acting as a component of the agentic AI system in accordance with an embodiment. In the Figures, reference signs can be omitted as is consistent with accepted engineering practice; however, a skilled person will understand that the illustrated components are understood in the context of the Figures as a whole, of the accompanying writings about such Figures, and of the embodiments of the claimed inventions. DETAILED DESCRIPTION OF THE DRAWINGS FIG. 1 depicts an agentic AI system architecture 100 able to respond to user requests made through a communication omnichannel 102 in accordance with an embodiment for the purposes of the present technology. The system architecture 100 includes a set of multiple agents 104, each agent being assigned to various function