ActionNex: A Virtual Outage Manager for Cloud Computing

TL;DR

ActionNex is a production-grade AI system designed to assist in outage management for cloud computing by providing real-time, context-aware recommendations based on multimodal operational data.

Contribution

It introduces a hierarchical memory and reasoning framework that integrates knowledge distillation, real-time signals, and human feedback for outage assistance.

Findings

Achieved 71.4% precision and 52.8-54.8% recall on real Azure outage data.

Successfully piloted in production with positive early feedback.

Supports end-to-end outage management with real-time updates and continual learning.

Abstract

Outage management in large-scale cloud operations remains heavily manual, requiring rapid triage, cross-team coordination, and experience-driven decisions under partial observability. We present \textbf{ActionNex}, a production-grade agentic system that supports end-to-end outage assistance, including real-time updates, knowledge distillation, and role- and stage-conditioned next-best action recommendations. ActionNex ingests multimodal operational signals (e.g., outage content, telemetry, and human communications) and compresses them into critical events that represent meaningful state transitions. It couples this perception layer with a hierarchical memory subsystem: long-term Key-Condition-Action (KCA) knowledge distilled from playbooks and historical executions, episodic memory of prior outages, and working memory of the live context. A reasoning agent aligns current critical events to preconditions, retrieves relevant memories, and generates actionable recommendations; executed human actions serve as an implicit feedback signal to enable continual self-evolution in a human-agent hybrid system. We evaluate ActionNex on eight real Azure outages (8M tokens, 4,000 critical events) using two complementary ground-truth action sets, achieving 71.4\% precision and 52.8-54.8\% recall. The system has been piloted in production and has received positive early feedback.