Artificial Intelligence at Barclays – Current Initiatives

Niccolo Mejia

Niccolo is a content writer and Junior Analyst at Emerj, developing both web content and helping with quantitative research. He holds a bachelor's degree in Writing, Literature, and Publishing from Emerson College.

Artificial Intelligence at Barclays - Current Initiatives

Barclays is a UK bank ranked 20th on S&P Global’s list of the top 100 banks. Like other top banks, Barclays has forayed into AI for a variety of use-cases. The bank seems to work with AI vendors more than it builds AI applications in-house, which aligns with the general trend of AI adoption in financial services: 68% of the AI products we researched as part of our AI Opportunity Landscape research in financial services were bought from vendors.

Our deeper analysis of AI at the top global banks and financial institutions, including banks like Citi and HSBC, can easily be discovered with Emerj Plus. The Emerj Plus matrix view allows members to quickly discover what top players in their industry are doing with artificial intelligence and which AI applications are driving the highest ROI across key areas in their industry.

In this article, we discuss three of Barclays’ AI initiatives in particular, as well as the AI vendors the bank claims to have worked with:

  • Risk Modeling with Simudyne: Predictive analytics for analyzing the risk of lending to a loan applicant.
  • Voice Recognition for Authentication with Nuance: Natural language processing-enabled identity verification and authentication via voice recognition.
  • Business Process Automation with IBM: Automated debit and credit card deactivation and customer feedback analysis.

We begin our analysis of Barclays’ AI initiatives with their “agent-based” modeling software, which they built with help from Simudyne.

Agent-Based Modeling – Simudyne

In July of 2019, Barclays announced a partnership with the AI startup Simudyne, which specializes in what the company calls “agent-based modelling.” This can be understood as the simulation of various banking situations using transactional data to build simulated customers within a bank’s ecosystem. This purportedly allows Barclays to simulate the banking and loan markets to produce detailed predictions.

Barclays claims that adopting Simudyne’s software helped them provide more robust offerings for their customers, such as new financial services.  This is likely because the software allows the client to develop a thorough understanding of the possibilities and risks associated with lending to a given customer or making a given investment.

In addition to gauging the risk associated with individual customers, Barclays employees can purportedly use Simudyne’s solution to:

  • Analyze mortgages: Simulate a relationship between customers and lenders by modeling housing and lending markets in conjunction with customer behaviors.
  • Stress Test: Users can scale up their simulation to represent a greater portion of the market, or scale it down to show granular details about specific situations. The user can test how accurate their simulation is by continually scaling it up until the system cannot populate any more results.
  • Contagion Risk: Create risk simulations in order to identify new investment risks and better understand how risky investment can lead to more precarious situations such as low liquidity.

Simudyne’s agent-based modelling software most likely runs on predictive analytics, an AI approach for making predictions based on historical data. For example, the bank can choose to see what might happen if they chose to lend to three customers with varying credit scores. When running this simulation, they can choose how much mock-data should be attached to each simulated customer.

This mock data can include information from past investments and loan applications. The software could then purportedly determine how likely each customer is to pay back their loan in full. 

The bank can also purportedly create a simulation of their investment applicants’ relationship to each other as well as other banks. Customers are organized according to the interactions and relationships they have with banks, credit unions, and other entities such as the IRS. The ML algorithm uses these factors to score customers on how likely they are to see profit from the investment or to pay their loan back. 

Barclays also claims to run Simudyne’s solution through the cloud, which allows them to create simulations that encompass an even wider subset of data. This is because Barclays can now access all of their enterprise data from any of their data science labs. This means they can likely simulate large shifts in banking trends and begin preparing for the changes most likely to happen ahead of time. 

For example, Barclays could see an influx of underserved loan applications over a short period of time. The bank’s risk managers could then run an agent-based simulation when trying to decide which of the applicants to lend to. This might show that nearly half of the applicants are indebted to other entities, and so are less likely to fully repay their loan with Barclays. 

Voice Recognition for Authentication – Nuance

Barclays also adopted a voice recognition solution from Nuance Communications to authenticate and verify its customers identities. Nuance refers to this capability as “voice biometrics,” and it runs on natural language processing (NLP) technology.

The bank claims its customer service agents are able to automatically detect if they are actually talking to who the customer claims to be or someone else and refuse access to those who do not pass authentication.

Nuance claims that their clients’ users can see the following benefits when using voice biometrics:

  • Improved customer experience: Customers may have less need to remember a password or PIN if they can be recognized with their voice.
  • Enhanced security: Fraud rates may decrease in frequency when authentication is unique to every individual as opposed to based on security questions that are easily compromised. 

Nuance claims the software detects the customer’s voice and authenticates it during the beginning of a customer service call without interrupting the call. If the software detects that the caller is not who they say they are, the customer service agent can choose to deny them access to personal information.

At this point, the agent may also be able to ask the customer to verify their identity through a password or PIN in order to continue receiving service.

The following video from Nuance explains how their biometric authentication capability works:

Nuance lists a case study on their website detailing Barclays’ success with its software. Barclays conducted client research and purportedly found that their customers were looking for a better user experience within the telephone channel security process. Additionally, the bank’s client service center relationship managers spoke up about how it is uncomfortable to ask their clients a comprehensive set of security quests after establishing a good rapport with them.

After implementing the voice biometrics solution from Nuance, Barclays saw a 90% reduction in complaints about the security service. They also saw a 15% reduction in average call times, allowing them to field more client queries in less time.

Business Process Automation – IBM

Barclays also adopted two solutions from IBM: Business Process Manager and Blueworks Live, a cloud-based storage system for documenting customer journeys as data.

Using Business Process Manager and Blueworks Live in conjunction, Barclays employees can purportedly:

  • Access customer journeys globally and determine how well a customer’s needs were satisfied.
  • Allow customers to receive SMS messages about attempted fraud with their account, which puts them in control of how they want to manage their funds should they be compromised.
  • Process customer service reports such as lost or stolen credit cards and possible identity theft.

The video below explains how Blueworks Live works:

IBM provides a case study which explains Barclays’ success with both the IBM Business Process Manager and Blueworks Live. The case study states that the solution helped Barclays gain a better understanding of how it documented customer journeys within the enterprise, how well they control them, and how customers feel about them.

The bank was purportedly able to take in reports of missing or compromised cards and replace them 67% faster. Additionally, they could also iterate on customer processes and roll out new ones 88% faster. 

Emerj for Financial Services Leaders

Leaders at large financial institutions use Emerj AI Opportunity Landscapes to discover how their competitors are leveraging AI to drive value in key areas like customer service and fraud. They use our research as a foundation for their own AI strategies, allowing them to continue to win market share well into the future. For more on how Emerj can help, learn about Emerj Research Services.


Header Image Credit: Barclays

Stay Ahead of the AI Curve

Discover the critical AI trends and applications that separate winners from losers in the future of business.

Sign up for the 'AI Advantage' newsletter: