Collaborative AI Agents and Critics for Fault Detection and Cause Analysis in Network Telemetry

TL;DR

This paper introduces a federated multi-agent system with collaborative AI agents and critics for fault detection and analysis across various domains, ensuring privacy and convergence guarantees.

Contribution

It develops algorithms for collaborative control of AI agents and critics with privacy-preserving, multi-modal fault detection and cause analysis, supported by convergence proofs.

Findings

Effective fault detection and cause analysis demonstrated in network telemetry.

Communication overhead scales with modalities, independent of agent count.

Convergence guarantees established for the collaborative algorithms.

Abstract

We develop algorithms for collaborative control of AI agents and critics in a multi-actor, multi-critic federated multi-agent system. Each AI agent and critic has access to classical machine learning or generative AI foundation models. The AI agents and critics collaborate with a central server to complete multimodal tasks such as fault detection, severity, and cause analysis in a network telemetry system, text-to-image generation, video generation, healthcare diagnostics from medical images and patient records, etcetera. The AI agents complete their tasks and send them to AI critics for evaluation. The critics then send feedback to agents to improve their responses. Collaboratively, they minimize the overall cost to the system with no inter-agent or inter-critic communication. AI agents and critics keep their cost functions or derivatives of cost functions private. Using multi-time scale stochastic approximation techniques, we provide convergence guarantees on the time-average active states of AI agents and critics. The communication overhead is a little on the system, of the order of $\mathcal{O}(m)$, for $m$ modalities and is independent of the number of AI agents and critics. Finally, we present an example of fault detection, severity, and cause analysis in network telemetry and thorough evaluation to check the algorithm's efficacy.