Aidan McEvoy is Vice President Issuer Modernisation at TSYS, the largest third-party payment processor for issuing banks in North America and one of the largest in Europe. In 2019, TSYS merged with Global Payments, an American multinational financial technology company. With a strong background in the cards and payments industry, Aidan is currently part of the business leadership on the TSYS’ modernisation programme, which includes the migration of legacy applications and technology services to the cloud. The use of artificial intelligence is being explored to test new software more efficiently and improve the end-user experience.
Why is it important for TSYS to move to a new cloud software system and why is testing essential?
Our modernisation programme aims to move a large, monolithic system – written primarily in COBOL and Assembler – to the cloud. When developing new applications, our technology team first puts them through internal testing cycles. Once these are complete, a team called IBA – Independent Business Assurance – takes over and conducts extensive quality assurance and user acceptance testing before the applications reach actual customers. In essence, IBA acts as an internal quasi-customer to ensure the highest quality before each release. The cloud project is divided into releases that focus on customer-facing applications and more complex releases that involve our back-end systems, where millions of accounts reside. This backend is a large mainframe-based system that has evolved over the last 30 years. Our role is to deconstruct and rewrite it in Java, migrate it to the cloud and test it thoroughly to ensure it meets our standards before going live.
How does AI fit into this process and what are the expected benefits?
The sheer size and complexity of the project makes traditional testing techniques impractical. We want to automate the creation of test cases, which will significantly speed up the test planning process. We’re exploring AI to distil multiple sources of information, such as user manuals, training materials and business process documents, to generate test cases and scenarios. The AI engine learns from this data and generates the necessary requirements, which we then use to create user stories, acceptance criteria and test cases. This approach aims to significantly reduce manual effort and manage complex functionality testing that would otherwise require extensive manual scripting. Although we’re still at an early proof-ofconcept stage, we see AI as a way to potentially save a lot of time and money by streamlining the testing process. We need to have a very critical testing regime because instead of using a lift and shift approach to move to the cloud, we’re rewriting the code for our legacy applications.
Are there potential cost and efficiency savings?
The complexity of testing in our operations is huge. For example, consider the interest calculations on a credit card – how a payment is applied and how transactions are processed. When we rewrite the code for such tasks, we need to make sure it works flawlessly. By using AI to automatically generate thousands of test scenarios, we can save a significant amount of manpower and time that would otherwise be spent writing these scenarios manually. We are excited about the potential of AI and we are actively looking at ways to integrate it into our processes.
“We are excited about the potential of AI and we are actively looking at ways to integrate it into our processes.“
Beyond testing, what are the broader implications of AI for your organisation?
AI will likely revolutionise our business. We’re exploring its applications right across our value chain. This includes App Dev, DevSecOps, Operations and Client Success Management teams. Our operations team is evaluating AI to analyse past problems and prevent future ones, while we’re also using AI to enhance our training programmes. A global AI centre of excellence has been established within Global Payments , marking the beginning of our journey. Our goal is to use AI to improve efficiency and quality, although we’re still in the early stages of implementation. There’s a lot of energy around global payments and the potential of AI right now. How easy will it be to integrate? We will have to wait and see. Full integration remains to be seen. We clearly see the potential of AI, but right now we are just scratching the surface.
This interview was taken from our recent report How AI is shaping the future of financial services.
