How automating insights can accelerate efficiencies in banking

By Fiona McNeill, Principal Product Marketing Manager, Red Hat, 

For financial services firms, the effort to do more with less seems like a never-ending goal. Now, with advanced digital technologies such as artificial intelligence (AI), machine learning (ML), cognitive computing, and other modern initiatives to streamline legacy processes and help reduce costs, the effort is gaining renewed ground.

Automation in banking began in earnest decades ago. In the early 1950s, Bank of America began working with Stanford Research Institute to develop a computer-based check processing system and machine-readable checks that would help the bank more efficiently handle the growing amount of paperwork involved in bookkeeping.

At the time, banks struggled to keep up with the flow of paper and the manual check-clearing processes involved with billions of checks being written each year. Engineers developed Electronic Recording Machine, Accounting (ERMA), and the M/CR, or magnetic-ink character recognition check coding system. Then in the late 1960s, automated teller machines (ATMs) came onto the scene, fundamentally impacting banking servicing.

Fast forward to today, and robotic process automation (RPA) is poised to digitally transform business processes in the banking and financial sector. RPA in its most general form is the application of software bots to perform tasks typically performed by humans. This includes processing a transaction, providing assistance, manipulating data, triggering responses, and communicating with other digital systems. While RPA is still a burgeoning technology, the RPA global market is expected to reach $3.11 billion by 2025, according to a study by Grand View Research, Inc[1].

RPA solutions can help customers digitally capture business logic and develop applications which automate business operational processes and decisions. The solutions integrate process management, rules management, resource optimization, and complex event processing (CEP) technologies into a single, integrated, open source platform. This can help simplify business operations, streamline tasks and reduce the time-consuming, error-prone manual data tasks, which in turn can increase speed and accuracy.

Even greater value to operational efficiencies might be had by embedding AI/ML and cognitive computing technologies into business processes – providing the capability to adapt to new events and goals – and such insights actioned based on decisioning defined to the system. Business operational efficiencies are no longer beholden to the limits of humans’ ability to do tasks, only to the limits of underlying processing power of the systems on which they run.

Transactional efficiencies could get a boost from advanced digital technologies, but there’s also the expectation that modernizing core banking systems will improve efficiency. More and more banks are evolving their legacy, closed systems to digital banking systems that use application programming interfaces (APIs) to create more flexible and streamlined operational and transactional environments.

By using open APIs, banks can more easily integrate data from disparate systems and sources. They may also positively impact business development by helping to accelerate partnerships and streamline partner integrations. The open nature of APIs can foster creativity and increase the rate of innovation to improve transactional insights leading to better actions.

No matter where the data comes from, strengthening your business operations environment relies on innovations, and an open source technology foundation provides a stable yet flexible platform that can scale and adapt so that customers receive a streamlined experience that meets their expectations.

Incumbent banks should know they have to modernize their organization to compete in a world where customers want better and more personalized digital experiences. Eager to realize the cost-savings and increased revenue that can result from micro-targeting products and services, they can adopt next-generation technologies to transform their businesses to lead their market.

Digital leaders are focused on end-to-end customer experiences. Processes, policies, and procedures defined for branch networks are being reimagined to support new digital customer engagement. By modernizing the back office and business processes, banks have an opportunity to streamline, codify, and thereby automate – which, in turn, can reduce friction caused by manual checks and inconsistent policies. This can enable more seamless customer experiences and speedier customer service, with transparency into servicing while reducing operational costs.

Artificial Intelligence (AI) is one of the leading digital technologies that’s captured the attention of financial services firms. While a number of use cases have emerged, one at the top of the list is its ability to help detect financial crime.

With increasing stores of event data, banks are challenged to analyze it given the old ways of storing, then analyzing data. Modern technology can help discover and predict anomalies in data without storing it first. Ultimately the goal is to do real-time detection as triage to help minimize the number of false positives investigated.

According to an article from Deloitte, it is the cognitive capabilities associated with machine learning and natural language processing that are expected to make fraud detection models more robust – stronger and more accurate. As described by the Cognitive Computing Consortium, by their very nature cognitive systems can be distinguished from other forms of AI in their ability to adapt and learn from iterative human interaction.

Ultimately, it is the results that matter, reduction in false positives of 95 percent to 50 percent, along with a reduction of 27 percent in manual effort have been cited in a case using modern machine learning techniques – helping discover the undefined unknowns in data. However, it remains to be seen how much better over time these systems will become if AI and cognitive systems come together, with experts who can label data and teach the algorithms iteratively, like that of machine learning techniques in which an algorithm seeks to maximize a value based on rewards received for being right.

We are seeing financial firms marry operational efficiency efforts with AI/machine learning/cognitive computing – creating an additional layer of automated insight that is designed to optimize bank service processing. Part of that optimization can also come from hybrid cloud adoption, in which AI and machine learning models are available to operational systems in the data center and/or in a public cloud.

Native cloud adoption can include the use of Linux containers containing the libraries, dependencies, and files teams need, and these containers can be spun up and down on-demand. Just imagine: analysts can define the rules that automatically execute business decisions, informed by insights from embedded algorithms. Those algorithms, in turn, are part of the pre-approved library defined by AI and domain experts. All of this could be from a self-service environment that doesn’t require your technology organization to spend time provisioning the tools, the data, or the processing capacity.

Of course, bringing these kinds of capabilities into new products beyond operations is within the realm of open banking. More banks seem to be realizing the value of co-creating products and services to expand their market reach to help them achieve new value streams. Combining back office operational efficiency and embedded intelligence with data sharing via open banking APIs should further propel digital leadership in financial services.

These technologies hold much promise, and banks should understand they need to rethink their technology investments to include them. But knowing what they need to do and figuring out how to do it can be two different things. Banks will have to be sure to aim and hit the digital high points that best fit with their long-term business plans aligned to customer journeys at the core. Today’s dynamic customer environment should only continue, with new entrants and new ways of providing banking services. Perhaps the most prudent strategy is to plan for change.

These technologies have one thing in common. A successful return on technology environments that are mutable to business needs often depend on a willingness by the firm (and its leaders) to accept the cultural, process, and policy metamorphosis necessary to make them – and the larger digital transformations they can facilitate – work. This is a culture change for a traditional long-standing industry.

It’s going to be challenging to digitally transform banks, yet a path must be chosen and navigated, all while the banking landscape continues to change.

[1] Robotic Process Automation Market Size Worth $10.7 Billion By 2027, Grand View Research 2020

[2] Why artificial intelligence is a game changer for risk management, Deloitte 2016