As fintech continues to deepen its footprint, the securities industry's digital transformation is entering its most demanding phase.
With cloud computing, microservices, and containerization becoming the norm for infrastructure, operations targets have expanded from traditional physical servers to virtual machines, containers, and multi-cloud resources. The scale and complexity of operations have grown exponentially,rendering fragmented, manual approaches unsustainable. For a brokerage that has just completed a merger, this challenge is multiplied many times over: how to fuse two IT systems that have operated independently for years—and how to seize the opportunity to re-architect the entire operations function—is a question management must answer.
As the partner building its integrated operations platform, CanWay worked with Guolian Minsheng Securities to drive a comprehensive transformation centered on architectural restructuring and operational modernization, covering both platform capabilities and team development.
Inside Guolian Minsheng: Fragmented Operations Struggling to Support the Post-Merger Business Landscape
The merger between Guolian Securities and Minsheng Securities was one of the first completed securities industry restructuring projects following the release of new regulatory framework. It has been recognized as a representative example of efficient execution and smooth integration within the securities sector.
The two organizations brought highly complementary strengths. Minsheng Securities had strong capabilities and experience in investment banking, while Guolian Securities had established advantages in wealth management and asset management. Geographically, one had a strong presence in central China, while the other maintained deep roots in the southern Jiangsu region.
Following the merger, the combined organization significantly expanded its business scale across wealth management, investment banking, asset management, and proprietary trading. However, this rapid business expansion also introduced higher requirements for IT stability, operational efficiency, and service management capabilities.
At the beginning of the project, CanWay conducted a comprehensive assessment together with the operations team of Guolian Minsheng Securities and identified four interconnected operational challenges.
First, fragmented configuration management. Two separate CMDB systems held asset data across five different data centers and hybrid cloud environments, with no unified inventory, no asset relationship model, and no reliable data foundation for impact analysis or change risk assessment.
Second, inconsistent process standards. The two companies maintained independent IT service management models, with incident management, change management, and resource request processes distributed across multiple systems. The absence of unified service catalogs and SLA governance increased collaboration complexity across teams.
Third, siloed monitoring and alerting. Guolian Minsheng relied on Zabbix and several other tools to collect monitoring data, resulting in fragmented alert sources and excessive redundant notifications.Without centralized alert aggregation, correlation analysis, and ticket integration, troubleshooting required engineers to switch between multiple systems during incidents, reducing response and recovery efficiency.
Fourth, fundamental architectural constraints. The legacy platform did not support containerized hybrid deployment or cross-data-center disaster recovery. It was also incompatible with domestic technology stacks, including Kunpeng and Hygon processors and Kylin operating systems, making it difficult to meet future requirements for technology independence and regulatory compliance.
These challenges were not independent issues but closely interconnected.Based on the assessment results, CanWay and Guolian Minsheng Securities reached a critical consensus: the post-merger operations system could not be improved through incremental tool stacking. A systematic transformation at the architectural level was required.
The transformation focused on three dimensions:
First, the operations platform technical architecture - establishing the technical framework and operating paradigm to support stable business services.
Second, the operations team - enabling the team to fully leverage new platform capabilities and continuously evolve operational practices.
Third, architecture governance and application lifecycle management — providing standardized and centralized governance perspective for management.
Co-Creating a Blueprint and Building a Unified Foundation
Instead of addressing individual issues separately, the project adopted a long-term transformation strategy centered on building a solid foundation, embedding AI capabilities throughout operations, and continuously evolving the platform.
The first step was to establish a integrated operations foundation by implementing four core capabilities: CMDB, ITSM, monitoring, and alert management. This enabled the consolidation of operational data and workflows while accumulating high-quality configuration and observability data to support future AI-driven scenarios.
The next phase focuses on applying AI Agents across scenarios such as inspection, diagnosis, self-healing, and change assistance, gradually enabling human-AI collaboration and extending toward FinOps-based cost optimization.
AI capabilities serve as a continuous evolution path throughout the transformation journey. By continuously incorporating advances in AI technologies, the operations platform can remain adaptable and aligned with future requirements.
At the platform level, based on the integrated architecture of CanWay platform, the solution was enhanced across two key dimensions: capability extensibility and operational resilience, ensuring both long-term evolution and secure operation.
The first layer focuses on extensibility and technology autonomy. Through deep integration of a unified operations architecture,OneAgent universal collection, API Gateway, Operations PaaS, and AIDev capabilities,
the platform enables flexible expansion of operational scenarios and supports independent capability evolution across different operational domains.
The second layer focuses on architectural stability and compliance. Through containerized deployment, full-stack compatibility with domestic technology ecosystems, and dual-data-center disaster recovery architecture, the platform achieves continuous availability, security, and compliance.
Project Focus: Building a Unified Operational Foundation
1. CMDB Configuration Management Center
Built on a four-stage maturity framework— Asset Inventory → Data Integration → Business Mapping → Digital Topology — forming the data foundation for new IT operations.
(1) Model standardization
Software and hardware configuration models were redesigned with standardized attributes, relationships, and data definitions, covering resources across IaaS, PaaS, and SaaS layers.
(2) Automated Discovery and Management
Through plugin-based integration with platforms such as container cloud environments, the system enables automated discovery and management of hardware, software, and cloud resources, achieving an automation coverage rate of over 60%.
(3) Data governance and operations
Data governance mechanisms were established through validation rules covering attribute completeness, isolated configuration items, and relationship integrity.
A data quality dashboard was implemented with clearly defined responsibilities among Configuration Managers, Data Owners, and Auditors, enabling continuous improvement and maintenance of configuration data quality.
(4) Application Architecture Design Capability
An advanced O32 business topology design capability was introduced to support visual application architecture modeling and comparison between designed and runtime states, providing a new approach to application architecture governance in the securities industry.
(5) Business value realization
Achieved 100% management coverage of core system resources across five data centers, providing reliable support for critical operational scenarios including:Incident impact analysis, Change risk assessment, Asset inventory management
2. ITSM Service Management Center: Rebuilt standardized operations workflow using five low-code engines (workflows / forms / rules / reports / views):
(1) Process Migration and Optimization
Core workflows, including resource requests, application onboarding and retirement, and change management, were redesigned and optimized to align with the post-merger organizational structure and data center resource landscape.
(2) Service Catalog Management
A hierarchical service catalog was established with user groups, service permissions, and SLA metrics configured to enable end-to-end ticket lifecycle tracking.
(3) ITSM-CMDB Integration and Closed-Loop Operations
Deep integration between ITSM and CMDB enables workflows to automatically retrieve configuration information during approval processes and update asset status after completion.
Monitoring alerts can automatically trigger incident tickets, creating a closed-loop operational process from detection to resolution.
(4) Multi-Channel Service Access
The platform integrates with enterprise collaboration tools and office workflow systems, enabling both desktop and mobile access to improve service processing efficiency.
3. Centralized Monitoring Center
A unified monitoring platform was established to consolidate monitoring standards and capabilities across the merged organization.
(1) Comprehensive Monitoring Coverage
The platform provides full-stack monitoring across five major categories: Infrastructure servers,
Middleware, Databases, Application processes, Synthetic transaction and port monitoring
(2) Visualized Operations Management
Business dashboards, resource overviews, and topology views were implemented to provide business-oriented visibility into monitoring coverage and service health.
(3) Flexible extensibility
The platform reserves capabilities for application tracing and log analytics, providing a foundation for future observability expansion.
4.Alert Management Platform
A full lifecycle alert management framework was established to address alert overload and improve operational efficiency.
(1) Multi-Source Alert Integration
The platform integrated six major categories of alert sources and supported alerts from more than 20 heterogeneous monitoring systems.
(2) Intelligent governance
Implemented alert enrichment, consolidation and merging, suppression, and topology-based correlation analysis, significantly reducing ineffective alerts.
(3) Closed-loop resolution
The platform supports: Automatic ticket generation from alerts, Automated remediation triggers, Intelligent assignment and routing. Key operational metrics, including MTTI, MTTA, and MTTR, were introduced to measure and continuously improve incident response and recovery efficiency.
(4) Normalized Operations
A biweekly alert review mechanism was established to analyze alert volumes and false-positive rates by business domain and technology provider, continuously improving monitoring strategies.
From Experience-Dependent to Capability-Driven
Building the platform is only half the transformation; the other half is the people. Around the platform's capabilities,CanWay helped Guolian Minsheng chart four interconnected career development paths for operations staff:
1. SRE (Site Reliability Engineer)
By leveraging monitoring and observability capabilities, the operations team can adopt engineering practices such as SLOs and error budgets to measure and manage system reliability.
This shifts operations from reactive incident response toward proactive reliability engineering.
2. Operations Development Engineer
Using low-code engines and Operations PaaS development tools, repetitive operational tasks can be transformed into reusable automation tools and self-service workflows.
3. AI Agent Development Engineer
By leveraging the future AI Agent development platform, operational knowledge and troubleshooting experience can be encapsulated into intelligent assistants, such as inspection assistants and alert analysis assistants.
Large language models will become digital collaborators for operations teams, supporting engineers in daily operational decision-making and execution.
4. FinOps Engineer
By leveraging comprehensive CMDB asset data together with capacity and cost information, the team can establish resource visibility, quota management, and optimization mechanisms across cloud and on-premises environments.
This enables IT operations to not only maintain system stability but also continuously contribute to cost optimization and business efficiency improvements.
These four development paths are not independent. Together, they enable the operations organization to evolve from a task-oriented support function into an engineering-driven, intelligent, and value-oriented capability center.
Exploring Application Architecture Governance: Managing Design, Deployment, and Runtime States
Application architecture governance remains an emerging discipline in the industry, with no universally recognized best practices established yet.
As an extension of the unified operations transformation, CanWay and Guolian Minsheng Securities are jointly exploring a new governance approach that connects application design, deployment, and runtime management.
While:
CMDB answers: “What resources exist and where are they located?”
ITSM answers: “How should operational processes be executed?”
Monitoring answers: “Is the system operating normally?”
Application architecture governance further addresses:
“How should the system be designed, where are services actually deployed, and does the runtime environment remain aligned with the intended architecture?”
Through the integration of design, deployment, and runtime states, application architecture can be governed throughout the entire lifecycle:
Design State: Establish business, application, and technology architecture models while defining dependencies between systems.
Deployment State: Map application components to actual infrastructure resources, including servers and cloud resources.
Runtime State: Continuously identify architectural deviations and potential risks by combining monitoring data, configuration information, and operational metrics.
Looking Ahead: Moving from Platform Integration Toward Intelligent Operations
According to the joint roadmap, the value of the platform will be continuously realized and measured through ongoing operations.
Following the roadmap of “building a strong foundation in the first year, embedding AI throughout operations, and continuously evolving capabilities,” the platform will integrate large language model capabilities into operational scenarios.
Future applications include: AI workflow assistants, Intelligent monitoring assistants, Intelligent inspection assistants, Alert analysis assistants
These capabilities will accelerate the transformation of IT operations from traditional support activities toward AI-driven digital operations and measurable business value creation.
