Key Highlights and PPTs from the 2026 XCOPS Intelligent Operations Conference – Guangzhou

On May 22, 2026 the XCOPS Intelligent Operations Conference was held in Guangzhou, focusing on the evolution of operations and databases powered by large AI models. Attendees examined architecture evolution, practical deployments, and the transformation of operational efficiency.

Opening Remarks – Jiang Chunyu, Director of Big Data & Intelligence, China Academy of Information and Communications Technology

Jiang highlighted that large‑model‑based operational agents are becoming the core of intelligent ops and that AI‑database integration is moving from concept to practice. He noted challenges such as technical breakthroughs and high industry adoption costs, and announced ongoing research and standard‑setting collaborations to support digital transformation.

AI Large Models in Intelligent Operations – Prof. Wang Wei, Fudan University

Wang explained how large models accelerate the path to L5 intelligent operations, dissecting key technologies like system state perception, knowledge acquisition, and complex task decision‑making. He identified current shortcomings—insufficient knowledge base accuracy, missing knowledge, and inaccurate root‑cause diagnosis—and advocated for generalized data accumulation and deeper model utilization.

AI‑Driven Operational Upgrade – Song Hui, General Manager, Xinju Network

Song argued that AI must focus on high‑frequency, high‑value scenarios such as alarm and fault analysis. Xinju’s solution combines large‑model and small‑model collaboration for second‑level fault localization, multi‑agent vertical layering, and expert skill‑based root‑cause identification, built on the ZnAiops platform that integrates data and intelligence middle‑layers to achieve a full‑process closed‑loop from monitoring to self‑healing.

From Tools to Systems – Wang Guansheng, Senior Ops Tech Director, Meitu

Wang described the transition of AI from a mere tool to foundational infrastructure, citing challenges like data and capability silos. He shared Meitu’s strategies for AI‑enabled efficiency gains and emphasized the need for “digital employees” that understand SOPs and carry work badges.

Observability in Core Trading Systems – Shen Bo, Core Observability Lead, Guotai Junan Securities

Shen outlined the high continuity requirements and complexity of securities core trading systems. He proposed a decoupled, standardized, intelligent, visualizable observability platform, detailing multi‑source data unification, end‑to‑end tracing, and an intelligent diagnosis engine, and summarized three key success factors: methodology, organizational collaboration, and technology selection.

Multi‑Cloud Neutral Architecture – Liu Xiangyang, Academician, European Academy of Sciences

Liu stressed the difficulty of building and managing multi‑cloud environments and advocated for a unified digital foundation. He described an AI‑powered compute platform, full‑stack monitoring, automated ops, and intelligent database management, offering concrete recommendations for manufacturing digital‑foundation projects.

Professional‑Grade Intelligent Agents – Zhu Sunwei, Technical Director, Migu Video

Zhu noted that generic intelligent agents still underperform in vertical scenarios. He identified challenges such as data‑to‑knowledge conversion and data accuracy, and analyzed three architectural traits—RAG, MCP, and SKILL. He emphasized the shift toward strategic evaluation infrastructures that drive roadmap decisions, predict capability limits, and set quality standards.

Featured Sessions

Zhang Yingying, Head of Intelligent Ops Algorithms, Alibaba Cloud – “AI‑Driven Intelligent Anomaly Handling: From Detection to Root‑Cause Localization”.

Gao Xuesong, AI Ops Manager, Ping An Life Insurance – “Financial‑Grade Intelligent Ops Migration: Full‑Stack Trust‑Zone and Cloud‑Native Performance”.

Ma Zhen, Senior Ops Manager, Sina Weibo – “AI Agent Collaboration in Ops at Sina Weibo”.

Liu Haoyang, Observability Expert, Volcano Engine – “Unified Observability for Agentic Applications with OpenClaw”.

Wang Qiong, Head of Cloud‑Native SRE, Baidu YY Live – “Breaking SRE Barriers with Agents in Container Cloud”.

Bang Xuedong, Ops Development Lead, Tencent Music – “AIOps in the ‘Lobster Tide’: Building an Intelligent Ops Ecosystem”.

AI‑Empowered Database Evolution

Speakers discussed how AI and large models are driving databases from traditional storage engines toward intelligent data services, covering topics such as large‑model‑based DB ops, vector search, financial‑grade DB transformation, and AI agents reshaping data service models.

Qu Zhe, Database Architect, JD Tech – “AI‑Driven Governance for Financial‑Grade Databases”.

Cai Dongzhe, Partner & CPO, NineData – “Balancing Security and Performance in AI‑Powered Coding”.

Liu Xiaoguo, Chief Evangelist, Elastic China – “Vector Search and AI Agents with Elasticsearch”.

Song Xin, Head of Intelligent Data Platform, Qunar – “From Data‑Ticket to Intelligent Data Retrieval: SQL Agent at Qunar”.

Wang Xue, Database Expert, Postal Savings Bank of China – “Massive‑Scale Financial DB Transformation: End‑to‑End Practice”.

Exhibition Highlights

The exhibition area was bustling, with partners showcasing core products and engaging in collaborative discussions.

Acknowledgments

The successful conclusion of the conference was thanks to the contributions of all speakers, partners, media, and staff.