一句话：提出记忆增强的生产级系统 ActionNex，用三类记忆在多模态运维数据上推荐云故障全生命周期的下一最佳行动。
大型云故障要求响应者在噪声、多模态且持续演化的运维数据中做出高风险决策；现有 AI 辅助多局限于分诊、排障和根因分析，难以支持生产环境中的全生命周期故障管理。ActionNex 将碎片化多模态信号转化为结构化运维上下文，并基于三类互补记忆给出状态感知的下一最佳行动推荐：以 Key–Condition–Action 单元编码的长期运维知识、历史故障案例构成的情节记忆，以及当前故障状态推导出的工作记忆。作者在 8 个真实生产故障、约 900 万 token 多模态数据上评估，使用 2 个留出故障和 6 个增量学习 episode，发现动作级召回随情节记忆积累从 73.8% 升至 79.6%、从 65.1% 升至 73.5%，精度大体稳定；检索策略可调节精度—召回权衡，结构化记忆优于仅提示方法并减少过度生成。10 个服务上的 30 天试点初步表明该系统在真实运维条件下具有实用性。
Major cloud outages require responders to make high-stakes decisions from noisy, multimodal, and continuously evolving operational data. While recent AI-assisted approaches help with tasks such as triage, troubleshooting, and root-cause analysis, they provide limited support for lifecycle-wide outage management in production environments. We present ActionNex, a production-grade, memory-augmented system that recommends state-aware next-best-actions throughout the outage lifecycle. ActionNex converts fragmented multimodal signals into structured operational context and grounds recommendation in three complementary memories: long-term operational knowledge encoded as Key–Condition–Action (KCA) units, episodic memory from historical outage cases, and working memory derived from the current outage state. We evaluate ActionNex on eight real production outages totaling approximately 9 million tokens of multimodal operational data. Using two held-out outages and six incremental learning episodes, we find that action-level recall improves consistently as episodic memory accumulates, rising from 73.8% to 79.6% on one outage and from 65.1% to 73.5% on the other, while precision remains broadly stable. Additional analyses show that retrieval strategy provides a practical precision–recall control knob, and that structured memory outperforms prompt-only alternatives while reducing over-generation. A 30-day pilot across 10 services provides early evidence of usefulness under live operational conditions. Together, these results indicate that memory-grounded, human-in-the-loop recommendation is a practical and effective design for production outage management.
@inproceedings{paperbot12471,
title = {ActionNex: A Production-Grade System for Next-Best-Action Recommendation in Cloud Outage Management},
author = {Zhenfeng Lin and Haoji Hu and Ming Hao and Xuchao Zhang and Yudi He and Oleg Kulygin and Hatay Tuna and Chetan Bansal and Ze Li and Murali Chintalapati and Sheila Jiang and Salman Zafar and Angie Anderson and Ryan Zhang},
booktitle = {ASE},
year = {2026},
doi = {10.1145/3832783.3834472}
}