HEnRY: A Multi-Agent System Framework for Multi-Domain Contexts

TL;DR

This work leverages existing research in MAS to introduce a new solution and covers two distinct research paths: the first focuses on the system architecture, and the second on the collaboration between agents.

Abstract

This project, named HEnRY, aims to introduce a Multi-Agent System (MAS) into Intesa Sanpaolo. The name HEnRY summarizes the project's core principles: the Hierarchical organization of agents in a layered structure for efficient resource management; Efficient optimization of resources and operations to enhance overall performance; Reactive ability of agents to quickly respond to environmental stimuli; and Yielding adaptability and flexibility of agents to handle unexpected situations. The discussion covers two distinct research paths: the first focuses on the system architecture, and the second on the collaboration between agents. This work is not limited to the specific structure of the Intesa Sanpaolo context; instead, it leverages existing research in MAS to introduce a new solution. Since Intesa Sanpaolo is organized according to a model that aligns with international corporate governance best practices, this approach could also be relevant to similar scenarios.

Figures (4)

• Figure 1: The image depicts how domains work within Intesa Sanpaolo. The user interacts with domains, each contain business processes and a knowledge base. These domains rely on IT services provided by specific technological domains, which are managed by domain experts to ensure proper operation and support of internal processes.
• Figure 2: The system includes an agent called digital twin, designed to be a comprehensive assistant for the user offering complete customization. The facilitator, being aware of all domains, collaborates with the digital twin to solve problems across various areas. The digital twin can integrate useful information for the facilitator when further details are required. The domain agents are distributed and aligning the company hierarchical structure with that of the multi-agent system, ensuring both maintainability and security. As concern security, the compartmentalization of roles is effective, allowing each agent to verify the conditions necessary to carry out an operation. All agents share a session data service, where web and agents track operations, while agora facilitates cross-domain collaboration through an ephemeral agent called the mediator, who ensures control and security during access to inter-domain services.
• Figure 3: The figure illustrates the following scenario: there are two domains—the HR domain and the CV domain. In the HR domain, we find regulations and available positions within the company. In the CV domain, we find candidate CVs for the company. In the multi-agent system, there is a digital twin of the user, a facilitator, and two agents that can operate within these domains. In this simulation, the agents can access the knowledge of their respective domains.
• Figure 4: The figure illustrates the following scenario: a mediator agent creates two agents to solve a problem. As described in section \ref{['fsm-mediator']}, the mediator follows different stages to solve the requested problem with the other agents.