In the realm of modern IT operations, the rapid advancement of cloud computing, microservices, and containerization technologies has led to an exponential increase in both system scale and complexity. Effectively achieving efficient, stable, and intelligent system management amidst dynamic, heterogeneous, and continuously evolving environments has emerged as a fundamental challenge in contemporary operations practice.

Against this backdrop, we propose MicroOps, a next-generation intelligent operations platform designed to address these emerging challenges. By introducing the “Micro-Operations” approach, MicroOps decomposes intricate operational workflows into fine-grained units and orchestrates them through automated tasks, enabling efficient, controllable support for daily system management and fault recovery. Built upon a cloud-native technology stack and deeply integrating techniques from Data Mining and Information Retrieval, MicroOps seamlessly combines automated operations (AIOps), workflow orchestration, real-time monitoring, intelligent alerting, and self-healing capabilities. This empowers enterprises to transition from reactive incident response to proactive, intelligent governance.

Our Group and MicroOps platform focus on three closely connected core directions to advance the development of a new generation of intelligent and automated operations systems:

1. In AIOps, we conduct in-depth research on intelligent data analysis, anomaly detection, and root cause localization. These efforts enhance incident response speed and system stability, enabling a shift from reactive alerting to proactive prevention.
2. ChatOps Framework Development: We are developing a ChatOps framework that integrates instant messaging tools with automated workflows. This allows operations teams to trigger and control tasks via natural language commands, reducing operational complexity while improving collaboration efficiency.
3. LLM-RAG Intelligent Assistants: We explore Large Language Model (LLM) and Retrieval-Augmented Generation (RAG)-based intelligent assistants for operations. By combining knowledge retrieval and language generation, these systems provide accurate, real-time, and context-aware decision support for operational tasks.

By integrating these research directions, MicroOps is committed to building an intelligent, adaptive, and human-centered operations platform. This foundation supports next-generation cloud-native infrastructures, intelligent monitoring systems, and self-healing architectures.

Ph.D Students

Master’s Students