In November 2025, the technical practice article titled Construction Practices of Yantai Bank's Integrated Operations Platform was published in the second half of the November issue of Financial Electronic magazine due to its technical foresight.
Yantai Bank has partnered with CanWay BlueWhale to successfully build an integrated O&M platform. The platform fully empowers the innovation of Yantai Bank’s O&M management and accelerates its digital transformation.
This article illustrates how Yantai Bank, through the development of an integrated operations platform and technological convergence, has established a next-generation operations system. This promotes the transformation of the operations model from decentralization to platformization, lays a solid foundation for intelligent operations, and achieves breakthroughts from system integration to capability enhancement. The initiative also empowers innovation in operations management at Yantai Bank and steadily advances the bank’s enterprise-wide digital transformation.
Under the tide of the digital economy, the banking industry is undergoing unprecedented technological and operational transformations. With the rapid iteration of technological applications and the exponential expansion of business scale, the traditional and fragmented "siloed" IT operations model can no longer meet the business demands of high availability and efficiency. In responese, Yantai Bank has proactively planned and established an integrated operations platform,aiming to build a centralized and platform-based operations management system. This initiative breaks down technical barriers, unifies data pipelines,and ultimately achieves comprehensive control over IT resources, service statuses, and operational processes.
During the development of the integrated operations platform, we came to realize that Intelligent Operations (AIOps), as a new-generation operations approach integrating big data and machine learning, has become a major trend in the global IT operations sector. In response, we drew insights from mature industry solutions such as AIOps and tailored them to the specific business characteristics of Yantai Bank, ultimately forging a practical and effective pathway suited to needs of Yantai bank.
Three Key Challenges: The Necessity of Integrated Operations
At the early stage of digital transformation, Yantai Bank’s IT operations faced three key challenges, which reflect the limitations of traditional operations models.
1.Fragmented and disorganized asset management.
Traditional IT asset management is resource-centric, with configuration information scattered across different tools and spreadsheets, leading to inconsistent and outdated data. In complex application architectures and dynamic cloud environments, operations teams struggled to form an application-centric holistic view of IT, resulting in low efficiency for change impact analysis and incident troubleshooting.
2.Blind spots and delays in status awareness.
Multiple monitoring tools generated massive alerts with inconsistent standards, creating “alert storms.” Data silos forced operations personnel to log into multiple platforms for analysis, resulting in low efficiency and a high risk of missing critical information. This prolongs the mean time to respond and recover (MTTR) and undermined business continuity.
3. Disconnected and bottlenecked process collaboration.
IT service processes—such as incident, problem, and change management—depend on offline communication and manual execution, leading to opaque procedures and inefficient cross-team collaboration. Fragmented operations lacked unified authorization, audit, and automation, increasing operational risks and limiting overall efficiency improvements.
Breaking Through: The Technical Architecture of the Integrated Operations Platform
To address these challenges, we started with top-level design and built an integrated platform that combines unified monitoring, centralized configuration management, online service processes, and automated operations. The core idea is to digitally reshape and platformize the three key elements of operations—objects, status, and actions—enabling deep integration of technology and management, and laying the foundation for intelligent advancement.
图1 烟台银行一体化运维平台技术架构
As illustrated in the technical architecture of the integrated operations platform, we divided its implementation into three stages: asset digitalization, action digitalization, and status digitalization, detailed as follows
1. Asset Digitalization: Building an Authoritative and Unique Configuration Management Database (CMDB)
Object digitalization is the cornerstone of integrated operations. Moving away from the traditional resource-centric approach, we built a “business application”-centric digital twin that reflects the logical relationships of business systems. To establish a unfied data model, we structured layered object models for infrastructure, PaaS platforms, and business applications, defining clear relationships to form an accurate IT resource topology network.
To address CMDB data drift, we shifted from a “passive data source” model to an “active digital twin.” By automated governance and continuous auditing, we integrated multi-source data through auto-discovery, agent collection, and API integration to automatically identify and correct anomalies. By introducing an operations knowledge graph and building on the CMDB, we visualize the complex relationships among configuration items — including physical connections and deployment dependencies — delivering explainable evidence for fault analysis and breaking through the "black-box" limitations of traditional AIOps.
2. Action Digitalization: Standardizing Processes and Enabling AI-Driven Dynamic Orchestration
Action digitalization aims to standardize, digitize, and automate operational tasks, thereby enhancing efficiency and controlling risks. We established unified IT service management center and automated operations platform. It solidified processes like incident, problem, and change management online. Using visual orchestration tools, standard operations are encapsulated into reusable atomic tasks and embedded into service workflows, achieving seamless integration between management and operational flows. At the intelligent stage, the goal is AI-driven automation and resource optimization. Key scenarios include:
Scenario 1: Intelligent Fault Self-Healing. Once the alert center identifies the fault source through intelligent root cause analysis, the AI diagnostic model automatically matches and triggers a predefined self-healing plan (e.g., service restart or resource scaling). This advances the response model from "human decision + automated execution" to "algorithmic decision + automated execution," significantly reducing the MTTR.
Scenario 2: AI-Driven Capacity Prediction and Dynamic Scheduling. Based on historical operational data, a deep learning time-series predication model is built to accurately predict resource usage. This model interacts with the elastic scaling mechanisms of the cloud-native platform to enable "predictive" dynamic scheduling, thereby improving overall resource utilization.
3. Status Digitalization: Building a Panoramic, Closed-Loop Intelligent Alert Center
Status digitalization focuses on accurate, real-time, and proactive perception of IT system states. A centralized alert processing center collects heterogeneous monitoring tool alerts via standardized APIs, and employs algorithms clean, deduplicate, consolidate, and suppress alerts, transforming “alert storms” into valuable events. Alerts are automatically correlated with CMDB configuration data to provide context and trigger automated ticket dispatch, creating a closed-loop process from detection to response. At the intelligent deepening stage, the alert center focuses on alert governance and intelligent root cause analysis. Machine learning algorithms aggregate and recognize patterns in raw alerts, converging them into key incidents. Multi-modal data fusion is applied for root cause localization, reducing fault identification from hours to minutes and achieving a performance leap from “alert collection” to “insightful fault detection.”
Achievement and Future Outlook
1.Summary of Phased Achievements
After more than a year of development and implementation, the integrated operations platform has begun to deliver tangible results. In terms of efficiency, automated operations have increased the efficiency of routine application releases and change deployments by over 60%, while standardized service process have reduced the average incident handling time by 40%. In terms of system stability, the application-centric CMDB and its associated topology have significantly shortened the time for fault impact analysis and localization. The unified alert center has reduced the mean time to detect (MTTD) major incidents by nearly 50%.
2. Future Outlook: Towards "Unmanned" Operations
The intelligent development of the integrated operations platform is an ongoing, evolutionary process. Looking ahead, we will deepen the application of operational data and incorporate AIOps technologies such as machine learning, knowledge graphs, and large language models (LLMs) to advance from "automated" to "intelligent" operations. AI will serve as an "augmented brain" and an "intelligent assistant" to human teams, becoming a powerful ally for the operations staff and driving the continuous evolution and innovation of the operational system. Our aspiration is to free experts from repetitive tasks, allowing them to focus on strategic innovation and solving complex challenges.
