How SmartStream derives wisdom from data noise

Banks and other financial institutions (FIs) are dealing with massive and unprecedented amounts of data, and it is only set to increase.

Andreas Berner, Chief Innovation Officer, SmartStream

These large data sets can expose financial institutions, especially those that rely on legacy systems and manual processes, to increased risk.

Artificial intelligence and machine learning (AI/ML) tools are essential if banks and financial institutions are going to face challenges when it comes to data, particularly surrounding data settlement.

Financial Technology Futures Talk with Andreas Burner, Chief Innovation Officer (CIO) at SmartStream Technologies, about how a company’s AI/machine learning technology can extract wisdom from data hype, bridge the knowledge gap created by employee disruption and how, in the age of machines, AI/can Machine learning frees humans to handle valuable qualitative data tasks.

While many financial institutions have a fairly high rate of automation within their current legacy systems, which means that they Can Processing large amounts of data, and sometimes the relevant information coming into these institutions is “not in perfect shape,” says Berner.

Legacy rules-based systems struggle with huge amounts of complex and messy data, and thus financial institutions need new ways to automate “not-so-beautiful data”, or else these companies become too reliant on manual labour. However, automating such unstructured data is very difficult.

“When we work with Level 1 organizations, we see this overreliance on manual processes in their exception management departments,” Berner says. “This is a huge pain point in the banks that we are currently targeting, and we are trying to make this process more efficient.”

big resignation

In the wake of the Covid-19 pandemic, many financial institutions are struggling with an influx of real-time digital payment data and an exodus of employees – often taking their accumulated wisdom with them.

However, there is wisdom to be gleaned from the data that financial institutions already own. Because financial institutions have to store a lot of this data for many years for regulatory reasons, “when we look at that data, it contains information about workflows and how employees make decisions in certain scenarios,” says Berner. This helps bridge the remaining knowledge gap when senior employees leave and a new group takes over.

If new employees can be mentored with software, it can actually help financial institutions prepare new employees, helping to pass on this accumulated wisdom.

Financial institutions are also experiencing a massive influx of low transaction value data thanks to the ubiquitous ubiquity of contactless digital payments. It is a trend that banks and financial institutions believe will increase.

When SmartStream works with banks on projects in the Innovation Lab, “we always hear that we have to be prepared to process larger amounts of data than we’re actually dealing with,” says Brenner.

Financial institutions, in the face of this trend, realize that thanks to players like Apple Pay and Google Pay, the amount of transaction data that needs to be settled will only increase.

“People pay far less digital transaction amounts than they would have in the past, so you see a large number of transactions, many of which are of very low value, and it is very difficult to reconcile them,” explains Berner.

However, managing these transactions is still important because if a large number of them fail – despite their low individual values ​​- in the aggregate, it is still a huge risk for any bank. “While contactless technology is great for customers, it can be a burden on their bank,” Berner adds.

Machine life

Artificial intelligence and machine learning are certainly key for financial institutions if they want to deal with the massive data sets that are now popular in finance. But in this age of machines, what role do humans play when it comes to managing and making use of financial data?

“Managing exceptions is the hardest thing to do in financial institutions because exceptions come in so many different flavors,” says Berner. “That’s why our latest technology, which we announce at Sibos, is AI for exception management.”

These exceptions can range from data problems to network problems, all the way to the vendor. With such a potentially diverse number of exceptions, management teams can be “huge,” says Berner, “because each problem is so specific.”

SmartStream addresses these issues with its proprietary ML techniques that can learn many of these issues and try to automate these issues. Humans act as moderators in this process, freeing them up to “focus on really difficult problems.”

The human element doesn’t have to wrestle with quantitative data—machines can handle that—but instead, people can shift their focus and attention to qualitative data problems, and improve these processes “dramatically,” adds Berner.

While the focus in SmartStream is exception management, where compromises are only one of the sources from which exceptions originate, “When we started talking to exception management departments, we discovered that we had solved a much bigger problem than just compromises,” Burner says. “Ultimately, it’s all about enhancing the customer experience.”

Exceptions often arise at a time when there may be a payment issue, a payment has been taken by a party that the customer does not recognize, or their account has been hacked. Artificial intelligence/machine learning techniques can solve these problems accurately and quickly.

“It would make a huge difference if the bank could solve this problem within half an hour or a week.”

Cracking open the black box

While the benefits of artificial intelligence/machine learning for banks and other financial institutions are numerous, many are still apprehensive about these emerging technologies to some extent. This is especially true when it comes to so-called black box technologies, where applications issue an answer or direction without revealing how they came to that specific conclusion.

Interestingly, SmartStream addresses this issue by explaining how these techniques arrive at their answer, and creating transparency by adding a window in the black box.

“We might say, ‘We’re suggesting this workflow to you because your data is in US dollars and there is a certain pattern in the data that will help customers,'” Berner says.

It not only reveals the “totals” of the black box, but also presents the data in a human-friendly manner. To do this, SmartStream has hired UI/UX teams to present the data in a way that is palatable and actionable to users.

“It’s not just about the data, but how do we display and present the data. We have huge amounts of information, but how do you present that information to the user in a useful way?”

Thanks to this investment in UI/UX, SmartStream was recently awarded a Red Dot Award, which caters to the communications and user experience design industry.

“We have invested a lot in the design of our app’s user interface, to make the UI and AI seamless, so that users feel supported and can trust the system,” Burner says.

Context is also key here, with only the exact information needed for the workflow displayed. “It’s very contextual, we’re showing the information we think will help the user, and nothing else.”

reinforcement

SmartStream has shown that when AI and machine learning are deployed in settlements, processing efficiency can be boosted by up to 20%. But there are other advantages, operational or otherwise, that a bank or financial institution can expect when these technologies are deployed.

“While we can automate 20%, we can also suggest to the user another 20% of workflows, and suggesting workflows to the user also increases efficiency significantly,” Burner says.

While this efficiency cannot be measured in addition to automation, “you can see that users are getting high quality data because the data has already been reduced before it is presented to them.” This arrangement, where humans and machines work together, has a significant positive operational impact.

The aforementioned employee momentum can also be improved, leading to efficiencies by relaying the employee knowledge accumulated during the onboarding process to new employees.

Learn by watching

The pace of innovation in this field continues. Artificial intelligence/machine learning technologies are getting smarter all the time. One new piece in SmartStream is Affinity, which uses observational learning to reduce the complexity of leveling.

Convergence is a component that can be deployed in the bank that works with the SmartStream application and observes user actions to understand the basis on which decisions are made.

“He’s constantly looking for patterns and when he finds patterns, he’ll remember what he learned,” Berner says.

Once activated, Affinity can be up and running right away, examining the vast amounts of data banks have to store and learning from past actions. “Therefore, there is no intensification period,” Berner explains. “We can train it using data from the past seven months, and even the past seven years.”

Affinity is fully integrated into SmartStream’s settlement solution, where it can automate workflow, suggest workflow, or measure workflow quality.

“It’s a very powerful component that we roll out with our software products.”


You can read in full The Daily News in Sibus The addon, sponsored by SmartStream, over here.

earlier this year, Financial Technology Futures I spoke with Jethro MacDonald,

Leave a Comment