Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Architecting Agentic Communities using Design Patterns

Zoran Milosevic, Fethi Rabhi

TL;DR

The paper addresses the challenge of engineering enterprise-ready agentic AI systems by introducing a three-tier taxonomy (LLM Agent, Agentic AI, Agentic Community) and a comprehensive design-pattern catalogue grounded in ISO OD P-EL formalism. It provides a systematic methodology (assess, compose, scope) to select and combine patterns, illustrated through a three-layer architecture (Foundation, Core Processing, External Integration) and a clinical trial matching case study that demonstrates governance, accountability, and runtime verification. The work contributes a taxonomy refinement, a 46-pattern catalogue with a unifying template, and formal semantics for Agentic Communities via deontic governance tokens and intent modeling, enabling provable safety and regulatory compliance in multi-participant, human-in-the-loop deployments. By integrating enterprise standards with formal verification, the approach promises reliable, auditable, and scalable production deployments in regulated industries such as healthcare. The framework thus bridges practical architectural guidance with rigorous formal semantics, supporting trust, governance, and adaptability as agentic AI technologies mature.

Abstract

The rapid evolution of Large Language Models (LLM) and subsequent Agentic AI technologies requires systematic architectural guidance for building sophisticated, production-grade systems. This paper presents an approach for architecting such systems using design patterns derived from enterprise distributed systems standards, formal methods, and industry practice. We classify these patterns into three tiers: LLM Agents (task-specific automation), Agentic AI (adaptive goal-seekers), and Agentic Communities (organizational frameworks where AI agents and human participants coordinate through formal roles, protocols, and governance structures). We focus on Agentic Communities - coordination frameworks encompassing LLM Agents, Agentic AI entities, and humans - most relevant for enterprise and industrial applications. Drawing on established coordination principles from distributed systems, we ground these patterns in a formal framework that specifies collaboration agreements where AI agents and humans fill roles within governed ecosystems. This approach provides both practical guidance and formal verification capabilities, enabling expression of organizational, legal, and ethical rules through accountability mechanisms that ensure operational and verifiable governance of inter-agent communication, negotiation, and intent modeling. We validate this framework through a clinical trial matching case study. Our goal is to provide actionable guidance to practitioners while maintaining the formal rigor essential for enterprise deployment in dynamic, multi-agent ecosystems.

Architecting Agentic Communities using Design Patterns

TL;DR

The paper addresses the challenge of engineering enterprise-ready agentic AI systems by introducing a three-tier taxonomy (LLM Agent, Agentic AI, Agentic Community) and a comprehensive design-pattern catalogue grounded in ISO OD P-EL formalism. It provides a systematic methodology (assess, compose, scope) to select and combine patterns, illustrated through a three-layer architecture (Foundation, Core Processing, External Integration) and a clinical trial matching case study that demonstrates governance, accountability, and runtime verification. The work contributes a taxonomy refinement, a 46-pattern catalogue with a unifying template, and formal semantics for Agentic Communities via deontic governance tokens and intent modeling, enabling provable safety and regulatory compliance in multi-participant, human-in-the-loop deployments. By integrating enterprise standards with formal verification, the approach promises reliable, auditable, and scalable production deployments in regulated industries such as healthcare. The framework thus bridges practical architectural guidance with rigorous formal semantics, supporting trust, governance, and adaptability as agentic AI technologies mature.

Abstract

The rapid evolution of Large Language Models (LLM) and subsequent Agentic AI technologies requires systematic architectural guidance for building sophisticated, production-grade systems. This paper presents an approach for architecting such systems using design patterns derived from enterprise distributed systems standards, formal methods, and industry practice. We classify these patterns into three tiers: LLM Agents (task-specific automation), Agentic AI (adaptive goal-seekers), and Agentic Communities (organizational frameworks where AI agents and human participants coordinate through formal roles, protocols, and governance structures). We focus on Agentic Communities - coordination frameworks encompassing LLM Agents, Agentic AI entities, and humans - most relevant for enterprise and industrial applications. Drawing on established coordination principles from distributed systems, we ground these patterns in a formal framework that specifies collaboration agreements where AI agents and humans fill roles within governed ecosystems. This approach provides both practical guidance and formal verification capabilities, enabling expression of organizational, legal, and ethical rules through accountability mechanisms that ensure operational and verifiable governance of inter-agent communication, negotiation, and intent modeling. We validate this framework through a clinical trial matching case study. Our goal is to provide actionable guidance to practitioners while maintaining the formal rigor essential for enterprise deployment in dynamic, multi-agent ecosystems.
Paper Structure (38 sections, 4 equations, 3 figures, 5 tables)

This paper contains 38 sections, 4 equations, 3 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Background
  3. LLM Agent, Agentic AI and Agentic Community
  4. Formal Foundation: ODP-EL Communities
  5. Contributions
  6. Design Pattern Catalogue
  7. Design Pattern Fundamentals
  8. Catalogue Overview
  9. Pattern Categories and Organization
  10. Agentic AI Architecture Design Methodology
  11. Overview
  12. Step 1: Assess Use Case Characteristics and Identify Candidate Patterns
  13. Step 2: Apply Pattern Composition Principles
  14. Step 3: Apply Scale-Appropriate Scoping and Implementation Planning
  15. Demonstration: Clinical Trial Matching Community
  16. ...and 23 more sections

Figures (3)

  • Figure 1: Three types of LLM-powered entities and their relationships. Agentic Communities coordinate heterogeneous participants including LLM Agents, Agentic AI systems, and human actors through structured protocols, shared infrastructure and agreed and computable governance specification.
  • Figure 2: Applying Design Patterns for Producing an Agentic AI Architecture
  • Figure 3: Pattern composition in clinical trial matching system demonstrating vertical composition (Layer 1 $\rightarrow$ 2 $\rightarrow$ 3), horizontal composition (patterns within layers), and cross-cutting composition (governance patterns spanning multiple layers).