Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Fetch.ai: An Architecture for Modern Multi-Agent Systems

Michael J. Wooldridge, Attila Bagoly, Jonathan J. Ward, Emanuele La Malfa, Gabriel Paludo Licks

TL;DR

The paper identifies a gap where modern LLM-driven agents overlook decades of MAS research, leading to centralized, brittle systems. It proposes Fetch.ai’s multi-layer architecture that fuses on-chain trust/discovery (Almanac, ANAME) and an economy (FET) with a development toolkit (uAgent), a deployment platform (Agentverse), and an intelligent orchestrator (ASI:One) to enable open, collaborative, economically sustainable MAS. The decentralized logistics use case demonstrates secure identity, protocol-driven interaction, and on-chain escrow, illustrating practical agent discovery, negotiation, and execution in a Web3-native environment. Collectively, the work offers a principled framework for building robust agent ecosystems that blend classical MAS practices with modern AI capabilities, advancing scalable, trustworthy, and decentralized multi-agent collaboration.

Abstract

Recent surges in LLM-driven intelligent systems largely overlook decades of foundational multi-agent systems (MAS) research, resulting in frameworks with critical limitations such as centralization and inadequate trust and communication protocols. This paper introduces the Fetch.ai architecture, an industrial-strength platform designed to bridge this gap by facilitating the integration of classical MAS principles with modern AI capabilities. We present a novel, multi-layered solution built on a decentralized foundation of on-chain blockchain services for verifiable identity, discovery, and transactions. This is complemented by a comprehensive development framework for creating secure, interoperable agents, a cloud-based platform for deployment, and an intelligent orchestration layer where an agent-native LLM translates high-level human goals into complex, multi-agent workflows. We demonstrate the deployed nature of this system through a decentralized logistics use case where autonomous agents dynamically discover, negotiate, and transact with one another securely. Ultimately, the Fetch.ai stack provides a principled architecture for moving beyond current agent implementations towards open, collaborative, and economically sustainable multi-agent ecosystems.

Fetch.ai: An Architecture for Modern Multi-Agent Systems

TL;DR

The paper identifies a gap where modern LLM-driven agents overlook decades of MAS research, leading to centralized, brittle systems. It proposes Fetch.ai’s multi-layer architecture that fuses on-chain trust/discovery (Almanac, ANAME) and an economy (FET) with a development toolkit (uAgent), a deployment platform (Agentverse), and an intelligent orchestrator (ASI:One) to enable open, collaborative, economically sustainable MAS. The decentralized logistics use case demonstrates secure identity, protocol-driven interaction, and on-chain escrow, illustrating practical agent discovery, negotiation, and execution in a Web3-native environment. Collectively, the work offers a principled framework for building robust agent ecosystems that blend classical MAS practices with modern AI capabilities, advancing scalable, trustworthy, and decentralized multi-agent collaboration.

Abstract

Recent surges in LLM-driven intelligent systems largely overlook decades of foundational multi-agent systems (MAS) research, resulting in frameworks with critical limitations such as centralization and inadequate trust and communication protocols. This paper introduces the Fetch.ai architecture, an industrial-strength platform designed to bridge this gap by facilitating the integration of classical MAS principles with modern AI capabilities. We present a novel, multi-layered solution built on a decentralized foundation of on-chain blockchain services for verifiable identity, discovery, and transactions. This is complemented by a comprehensive development framework for creating secure, interoperable agents, a cloud-based platform for deployment, and an intelligent orchestration layer where an agent-native LLM translates high-level human goals into complex, multi-agent workflows. We demonstrate the deployed nature of this system through a decentralized logistics use case where autonomous agents dynamically discover, negotiate, and transact with one another securely. Ultimately, the Fetch.ai stack provides a principled architecture for moving beyond current agent implementations towards open, collaborative, and economically sustainable multi-agent ecosystems.
Paper Structure (27 sections, 7 figures)

This paper contains 27 sections, 7 figures.

Table of Contents

  1. Introduction
  2. Agents: The Origin Story
  3. Multi-Agent Systems
  4. The Emergence of LLMs…
  5. … and LLM-based Agents
  6. Applications of LLM-based Agents
  7. LLM-based multi-agent systems.
  8. Multi-agent problem solving.
  9. Multi-agent systems for world simulation.
  10. Limitations of Recent Agent Frameworks
  11. An Architecture for Modern Multi-Agent Systems
  12. A Foundational Layer: On-Chain Trust, Discovery, and Economy
  13. The Development Layer: Core Components for Agents
  14. Event-Driven and Asynchronous Architecture
  15. Standardized and Stateful Communication
  16. ...and 12 more sections

Figures (7)

  • Figure 1: The standard model. We build an agent to operate on our behalf in some environment---autonomously pursuing our delegated goals/objectives.
  • Figure 2: A classic agent architecture, involving a "vertical decomposition" of intelligence into its component pieces: reasoning, planning, learning, etc.
  • Figure 3: Can LLMs provide the "brain" for agents?
  • Figure 4: Architecture layers and components.
  • Figure 5: Agent Registration, Search, and Message Exchange.
  • ...and 2 more figures