Liu Yu: Strengthening Financial System Quality Assurance Through Data Validation and Risk Identification

Before a system is officially put into operation, much of the work involved remains largely unseen: verifying whether functions perform as expected, ensuring that data flows accurately, checking whether different systems can connect smoothly, and assessing system stability under abnormal conditions. For Liu Yu, these seemingly routine checks have become an integral part of her professional experience. Since joining the headquarters of a commercial bank as a Technical Test Manager in 2021, she has participated in the technical testing of more than 20 financial system projects. Her responsibilities have expanded from functional, interface, and performance testing to automated testing, data validation, and the application of emerging technologies such as artificial intelligence.

Rather than simply carrying out assigned testing tasks, Liu places greater emphasis on making the testing process comprehensive and systematic. When approaching a new project, she typically begins by reviewing requirements, mapping business scenarios and system relationships, and then developing testing strategies and implementation plans. Throughout the project, she follows up on risks, defects, and progress. In projects involving information technology innovation, the migration of core host systems, and digital transformation, she has focused not only on whether functions are implemented correctly, but also on performance, data migration, interface and file delivery, and fault-injection scenarios. In one project, she participated in dividing a large volume of retail customer data into 2,458 equivalence classes and used traffic replay to validate 16.24 million production transactions, identifying and helping resolve more than 1,000 issues.

As testing scenarios have become increasingly complex, Liu has also explored ways to improve efficiency through technology. Automated testing was one of the areas she began working with relatively early. She has used scripts to replace some repetitive manual operations and expanded automation applications according to specific testing scenarios. In recent years, she has also explored the use of artificial intelligence and large language models to assist in generating test outlines and test cases. In combination with AI coding assistants, she has analyzed code functionality and test coverage, while also exploring the use of multi-agent systems to support test-case execution. At the same time, she has investigated the application of vision-language models (VLMs) and optical character recognition (OCR) technologies in multimodal user-interface testing. For Liu, the value of emerging technologies does not lie in the technology itself, but in whether it can address practical problems in real testing environments.

Methods such as traffic replay, fault injection, and performance testing have gradually become part of her practical toolkit. By simulating high-load conditions or abnormal states, testers can observe how systems perform under different circumstances and identify potential risks before they affect production. Liu participated in performance testing covering 37 systems and more than 200 transaction scenarios, helping drive the resolution of issues identified during testing. Such work requires patience as well as a strong understanding of the relationship between business processes and technical systems, since a seemingly localized problem may ultimately involve interfaces, data, or an entire business workflow.

In data-oriented projects involving management accounting, open banking, intelligent risk control, and anti-money laundering, Liu has participated in functional, interface, performance, and batch-processing validation, while also contributing to the improvement of test cases and data-quality assessment mechanisms. Compared with simply checking whether a page functions properly, data-driven systems require closer attention to the integrity of the entire data lifecycle—from generation and transmission to final presentation. Through years of practice, Liu has developed a clear working approach: first identifying data sources and interface files, then conducting targeted testing and validation. Rather than waiting for problems to emerge, she seeks to cover testing scenarios as comprehensively as possible and reduce the risk of defects escaping into later stages.

As a Technical Test Manager, Liu's responsibilities extend beyond technical execution. Requirements reviews, testing strategy development, risk assessment, schedule management, defect tracking, and production acceptance all require continuous communication with business teams, developers, data centers, branches, and external partners. She has participated in testing related to projects such as “Cloud Banking,” carrying out joint testing and performance validation across multiple channels and systems. Her long-term project experience has also strengthened her focus on testing standards and the accumulation of practical knowledge. Through mechanisms such as testing exit criteria and case repositories, she seeks to ensure that lessons and experience gained from individual projects can continue to support subsequent work

Liu's professional development did not begin entirely within the technology field. During her undergraduate studies, she studied management and later continued her education in accounting. After entering the financial industry, she gradually moved toward financial technology and technical testing. Her experience across different fields has influenced the way she approaches system problems: in addition to technical details, she also considers the underlying business logic and relationships within the data. Outside of work, she has received a first prize in a youth debate competition and participated in volunteer services for events including the China International Import Expo and the World Artificial Intelligence Conference. These experiences have also contributed to a broader professional perspective.

For Liu, testing may not be the most visible part of a financial technology project, but it requires continuous learning, careful observation, and patience. From functional validation and automation to exploring artificial intelligence and other emerging technologies, she continues to look for more appropriate ways to break down complex problems and identify potential risks as early as possible. Rather than pursuing new technologies for their own sake, she focuses on whether a particular method can solve a real problem in practice. This process of continuous learning, adjustment, and refinement has gradually become part of her professional approach after years of working in technical testing.