Member Content
Published
Share
Share on facebook
Share on twitter
Share on linkedin
Share on email

UiPath: Key Trends Shaping the Use of AI in Banking and Financial Services

by Nitin Purwar

Despite the setbacks of the past year, companies are deploying artificial intelligence (AI) for transformations that create a path for stronger growth. This is what we constantly hear in our conversations with bank customers around the world. While 2020 was challenging for businesses across the globe, banks relied heavily on robotic process automation (RPA) to improve customer service, streamline internal operations, and boost profits.

Banks were early adopters of RPA. Now they are at the forefront of adding AI built into the UiPath platform to unlock new possibilities and expand automation into all sorts of new areas. It’s a clear trend—85% of financial services organizations are currently using AI and 77% of senior executives anticipate AI to have high or very high business importance in the next two years.

We’ve seen companies embracing the broad, flexible AI technology provided by UiPath. And not just for well-known, RPA-specific tasks, but for innovative intelligent solutions that address some of the industry’s most pressing challenges.

How banking organizations use AI

There are several important areas where banks are using UiPath AI technology to innovate. These include:

  • Improving customer experiences. Companies want to create best-in-class digital client journeys through tighter integrations between back-end and customer-facing operations. Additionally, banks are rapidly digitizing all services, from retail offerings, payment platforms, to wealth and investment management journeys. This helps them keep pace with fintech startups that have bypassed older, legacy processes.
  • Shifting operating models. Physical branches are being displaced by virtual channels, such as chat apps, mobile banking, contact centers, email communications, and electronic catalogs. This has led to the introduction of intelligence into these channels to ensure smoother customer service.
  • Accelerated use of AI and analytics. Banks have realized the goldmine of data they have in their systems and want platforms that can help them use this data to their advantage. As part of this effort, they not only building their own machine learning (ML) models but also want out-of-the-box ML offerings that can be seamlessly integrated into their ecosystem. This includes scenarios like handling unstructured documents, defaults and forecast predictions, segmentation, and classification.
  • Risk mitigation. Using AI and RPA, banks are creating solutions that reduce or eliminate human error in many processes. That, in turn, helps reduce security issues that can result from those errors.

 

AI is embedded throughout the UiPath Platform. Banks are leveraging this AI functionality for a wide range of tasks. For example, discovering the right process candidates for automation using UiPath Process Mining. They can intelligently classify and extract data from documents with UiPath Document Understanding. Banks can also enable software robots to see and understand computer screens with UiPath Computer Vision. And they can consume and scale ML models using UiPath AI Center.

Banks are also benefitting from pre-trained, out-of-the-box ML models from UiPath, which can be quickly retrained to fit specific needs and scenarios. Let’s look at a few real-world use cases where results from these out-of-the-box UiPath models have created a big impact on customer and employee journeys.

Dramatic improvements to customer service through AI

Email as a channel for customer service has seen a huge uptick during the COVID pandemic, with volumes more than doubling for many banks. A leading global bank receives more than 1 million emails annually from customers making requests, asking questions, and filing complaints. In the past, managing these unstructured emails was a huge time and financial burden for the bank because the content was routed to various teams using manually intensive, rule-based keyword classifications. It also created bottlenecks that affected the bank’s responsiveness and quality of service to its customers.

The bank retrained the UiPath Text Classification ML Model to recognize the bank’s specific product and sub-product names and categories in unstructured email texts. UiPath robots were deployed with the newly trained model to automate the receipt, categorization, and automatic responses to select emails.

After deploying the AI solution, the bank saw a 90% reduction in average handling times, more than 93% accuracy in automated routing of emails and—most importantly—faster and better response times to customer queries. See more here: UiPath AI Center: Automating Complaint Classification Process – YouTube.
Using AI to efficiently handle millions of documents

The handling of unstructured documents—whether it is ‘know your customers’ (KYC) documents, credit agreements, policy documents, or trade agreements—has been one of the biggest challenges for banks. UiPath made it easier for them by providing models like the Document Understanding ML models. These and other ML models, like the Named Entity Recognition model and Question and Answer model, can be easily applied to fit specific customer scenarios. The implementation of these models took just a few weeks for some of our major banking customers.

One such example includes a major United States (U.S.) lending institution. The lending institution wanted to streamline the management of confirmation of swap agreements—the process in which one party exchanges the value or cash flow of one asset, such as stocks, for another. The bank was handling millions of swaps annually and sought a solution to automate and speed up the swap process.

The organization retrained the Document Understanding ML model to recognize key trade details. Then, using robots, the trades were reconciled against internal booking rules. Exceptions were automatically escalated to an operations team for review.

By using the UiPath Document Understanding ML model, the organization saw a 90% reduction in errors and realized an annual savings of about $1.2 million.
Expediting trade transaction processing using AI

In this example, a global bank wanted to improve how it handled tens of thousands of emails containing information for securities trade transactions. The unstructured emails were complex and required extraction of different trade fields and entity types. A single transaction could take as much as 8-10 minutes for a company employee to review and process, and there were voluminous errors.

The bank used the UiPath Named Entity Recognition ML model to recognize the various fields and entity types. Robots armed with the information from this model dramatically improved the transactions management.

Following deployment of the solution, the bank achieved a 95% increase in reduction in the average trade transaction handling times, along with a 92% accuracy rate for the information extraction.

The bank is now looking at using the same process for contracts, policy documents, and customer queries.

Watch the recording of the “AI in Banking and Financial Services” session of the UiPath AI Summit HERE.

Using AI to expand the impact of automation

Banks have long been pioneers in using technology to advance their businesses and continually provide customers with better service and products. UiPath enables bank customers across businesses and functions through a combination of RPA and AI technologies.

Today, we’re seeing a new wave of innovation by our bank customers. Much of this is driven by necessity. Customers have been using intelligent document processing capabilities from UiPath in the areas of KYC compliance, mortgages, and trade finance. UiPath pretrained ML models are being used in multiple areas including accounts payable, agreements and contracts management, retail and cards disputes management. Other common examples include loan default management, customer insights and next best offer predictions, trade supply chain management, and negative news screening.

Learn more about AI use cases in banking and financial services.

There are still difficult and sometimes unique challenges facing financial institutions due to factors such as regulatory mandates, the presence of huge legacy IT systems and processes, the complexity of the global economy, and customers’ expectations for better service and products.

Our bank customers are responding to these challenges by building solutions created with our comprehensive AI-enabled platform. They are benefiting from UiPath AI Center, Document Understanding, AI Computer Vision, and other products spanning the UiPath Platform. In this way, banks and financial institutions are not only tackling the difficult challenges of the past year but are putting themselves in even stronger positions for future growth.

To try out UiPath AI solutions for your business scenarios, sign up for the Enterprise Trial!

This post was co-authored by Amit Kumar, Global Banking Practice Lead at UiPath.

Read the original article here. Find out more about UiPath here.

Share this Article
Share on facebook
Share on twitter
Share on linkedin
Share on email
Related Insights
Featured
Amsterdam Fintech Week
Get ready for XFW 2021 coming up from the 4th-11th June!
AMLD5 Guide
A source for consulting PSD2 legislation coupled with commentary, tips & tricks, and applicability.