Valentyn Kropov, Chief Technology Officer at N-iX—a trusted software development partner for global businesses and Fortune 500 leaders.
The banking industry is currently experiencing a profound transformation fueled by technological advancements and the rapid growth of digital transactions. To remain competitive, companies are looking for ways to process more transactions faster and, at the same time, keep the core element of transactions—security and compliance.
Traditional technologies and algorithms are not able to cope with exponentially growing data to support these functions, and this is what pushes financial companies to use machine learning (ML) and advanced analytics (AA). Moreover, the emergence of AI-powered decision-making presents both opportunities and challenges for the future of the payment industry.
Recently, for example, my team worked with a leading U.K. challenger bank that offers digital banking services for small businesses. We have built a machine learning-powered decision engine that decreases payment processing expectation time from five minutes to under a second. The technology also provides real-time fraud detection and transparency of decisions and serves as an integration point for various internal bank systems.
In this article, I tap into insights gained from the case and explore the key four steps when developing innovative and safe payment solutions in the banking, fintech and retail industries.
1. Expand your view on users.
It all starts long before AI and machine learning; it starts with users and customer needs. To create effective solutions, financial institutions must conduct comprehensive user research that extends beyond just customers. Talk to as many users as possible, both internal and external, directly and indirectly.
For example, by involving bank employees and support staff in the research process, institutions can gain valuable insights into their unique needs and perspectives. Inclusive user research enables better decision-making.
2. Be ready to scale up or down quickly.
The digital era brings exponential growth, making scalability a critical consideration. Failure to account for scale can result in wasted infrastructure costs and missed opportunities.
Take the example of the 2018 Amazon Prime Day, where technical difficulties caused by an overloaded infrastructure led to an estimated loss of up to $100 million in potential sales. To avoid similar setbacks, financial institutions must invest in scalable solutions that can handle increasing transaction volumes and growing customer demands.
Another example that aligns with the concept of flexible and scalable solutions is the case of the London subway system, which has adapted to meet the evolving needs of its customers. In October 2022, Transport for London (TfL) released a press release stating that pay-as-you-go with mobile on the tube has become more popular than before the pandemic. The analysis conducted by TfL revealed that commuters are increasingly opting for the convenience and flexibility of pay-as-you-go transactions using their mobile devices, indicating a shift in customer behavior.
By implementing a pay-as-you-go system and incorporating mobile technology, TfL has embraced a scalable approach that allows commuters to make individual transactions for each journey rather than relying on monthly top-ups.
By ensuring scalability in advance, banks can optimize resource allocation and adapt seamlessly to the evolving digital landscape.
3. Stop undermining your processes with poor data quality.
The foundation of advanced analytics (AA) and machine learning (ML) technologies lies in data. Financial institutions possess vast amounts of customer data, and leveraging this information through AA is key to making informed decisions.
McKinsey’s report on AI-powered decision-making for the banks of the future emphasizes the importance of AA in transforming the banking industry. By employing sophisticated data analysis techniques, banks can uncover valuable patterns, trends and insights that can be used to offer personalized experiences, targeted marketing campaigns and efficient risk-management strategies.
Addressing data quality concerns, as highlighted by Gartner, ensures that the data on which decision-making relies is accurate and reliable, mitigating the risk of flawed outcomes. Robust data governance practices, data quality monitoring systems and data cleansing and enrichment tools are essential for ensuring accuracy and reliability.
4. Take a broader look at organization’s efficiency.
The integration of ML technologies offers an opportunity for banks to revamp and optimize existing processes. From manual operations to full automation, financial institutions can streamline workflows and eliminate errors.
However, it is also crucial to strike a balance between technological innovation and human oversight. It is essential to recognize that while technology plays a pivotal role, humans are ultimately responsible for its ethical and responsible use. Striking the right balance between innovation and oversight is paramount for creating a future in banking where customers’ best interests are safeguarded.
Safety Of AI-Powered Decision Making: Questions And Challenges
While AI-powered decision making holds tremendous potential for the payment industry, there are discussions about whether it is safe to rely on AI in payment processes.
One of the significant challenges in AI-powered decision making is the potential for bias in algorithms. Biased algorithms can lead to discriminatory outcomes, perpetuating inequalities and adversely impacting certain customer groups. Financial institutions must prioritize fairness and transparency in their AI models by regularly auditing algorithms, identifying and mitigating biases and involving diverse teams in the development and testing processes. As stated in Harvard Business Review’s article, “What Do We Do About the Biases in AI?,” “AI can help humans with bias—but only if humans are working together to tackle bias in AI.”
Another matter is that AI models need ongoing monitoring and evaluation to detect any anomalies or deviations from expected behavior. Financial institutions should establish comprehensive monitoring mechanisms to identify potential issues promptly. Regular audits and assessments of AI systems’ performance can help ensure they operate as intended and maintain the highest standards of safety and reliability.
To conclude, through thoughtful implementation and responsible usage, AA, ML and AI can shape a future where banks truly understand their customers, adapt to their evolving needs and drive sustainable growth in the dynamic world of finance.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Read the full article here
