Robotic process automation, or RPA, is a technology used across multiple industries to automate business processes. RPA software involves what are known as “software robots” to handle repetitive tasks traditionally handled by human employees. That said, there are no actual robots involved in the way one might see in manufacturing or heavy industry.
Throughout this article, we use the term “software robots” to refer to the RPA software itself. This particular report covers RPA applications in finance. Specifically, the vendors covered in this report offer RPA software and have made press releases claiming their software in some part makes use of AI. All of these vendors offer RPA software to enterprises.
The vendors covered in this report claim their RPA applications can accomplish the following tasks:
- Verification and Task Organization
- Credit Scoring and Fraud Detection
For further information on how applications such as AI and RPA can help banks and financial instituitions, our readers can download the Executive Brief for our AI in Banking Vendor Scorecard and Capability Map report.
Verification and Task Organization
Data scientists from Kofax work alongside the bank’s team to integrate their software into the bank’s existing data infrastructure and create software robots that can automate tasks such as data verification. Then, Kapow can collect the required information for each task and populate a report or an excel sheet with organized data as per the task requirements. This data is presented to human analysts who can then focus on reviewing the priority cases from the software output, thus reducing the time taken for the process in total.
Kofax recently announced that the latest version of their Kapow software also leverages AI and machine learning for automated document capture and desktop virtualization. Kofax launched what they call an intelligent screen automation for remote desktop environments. Users can use Kapow to speed up the development of software robots by using a new AI tool that allows the software to automatically identify user interface elements, such as labels and buttons.
Below is a short 4-minute video demonstrating how Kofax Kapow works:
Kofax claims to have helped an unnamed European bank automate parts of their verification process for Customer Due Diligence (CDD) and Know-Your-Customer (KYC) operations. The bank needed to comply with new European regulations regarding anti-money laundering. Due to the scale involved, the bank was finding manual operation of this process time and cost intensive.
These capabilities may prove to be more important to banks and financial institutions moving forward. Even though RPA applications are fundamentally different from formal AI applications powered by ML, this may be able to mimic a critical AI capability after adoption. We spoke to Yinglian Xie, CTO of Kasisto about how to use data and data science to reduce customer friction on our podcast, AI in Banking.
Customers may not feel appropriately satisfied if they have to wait a long time while their bank or chosen financial institution is verifying all of their information. When asked about which AI capabilities she thought were critical for the banking space, Xie said,
I think for banking there are definitely a lot of places where AI can play a big role, and for one thing, we all know the banking industry and we have the big data infrastructure ready. We actually have large amounts of data everywhere. And making sense of [it] to generate value for the banking business, as well as leveraging those businesses to make consumer experience a lot better … is definitely an area where ML and AI can help significantly.
The press release from UiPath also states that a spokesperson for the European bank said:
Highly skilled analysts were spending hours of their day combing through both internal and external systems to gather the relevant information on customers—a tedious exercise. And with legislation growing increasingly complex, we were constantly scaling up resources to keep pace with new requirements. Our manual approach was quickly becoming unsustainable, and we were in urgent need of a more automated and efficient way to manage CDD and KYC investigations.
The bank reportedly deployed the Kofax Kapow RPA solution to help automate the customer verification processes for CDD and KYC. The software robots comb through the bank’s internal customer data records, such as transaction histories, and external sources, such as Chamber of Commerce business registers, social media, news archives, and politically exposed persons databases to identify any suspicious history of illegal transactions in a customer’s data.
The bank spokesperson added:
We run the Kapow robots overnight; they gather any relevant information on the customers we have identified and put it all together in a report. When our analysts arrive at work the next morning, they have all the supporting data they need to inform CDD and KYC.
According to Kofax, the bank cut down on the time it took to collect data for CDD and KYC verification across its retail and corporate banking sectors from 15 minutes to 90 seconds and in from 10 minutes to 70 seconds respectively.
Kofax also lists Goldman Sachs, ING, American Express, Citi, HSBC, VISA, Wells Fargo, and J.P Morgan as some of their past clients.
We were unable to find any C-level executives with AI experience on the company’s team.
Credit Scoring and Fraud Detection
UiPath is a New York-based company with 1196 employees. The company offers a software called UiPath Enterprise RPA Platform, which the company claims can help finance businesses build and automate credit scoring, fraud detection, and other processes using robotic process automation.
The Enterprise RPA platform consists of three software components: Uipath Studio, Uipath Orchestrator, and software robots. UiPath claims banks and financial institutions can upload information such as customer credit histories to the software in the form of Excel sheets. Users can then create automation processes on the UiPath Studio using visual flowcharts. The studio software then allows users to publish their automation project to UiPath Orchestrator.
Then, Orchestrator enables the management, logging, and reporting of all the robotic processes published from UiPath Studio. This server software allows banks and financial institutions to select, run, and monitor the performance of each of their automations. For instance, a bank might automate the verification processes for retail credit scoring or fraud detection.
The banks can create robots with the platform that might gather customer information from several databases or comb through transaction data to identify suspicious activity. The system then provides detailed reports of the findings with all the data collected in one place. These reports can be analyzed by human employees. UiPath claims banks can reduce the time and cost required in these processes using their platform.
The UiPath Enterprise RPA Platform also allows users to built out additional capabilities for the software, such as intelligent optical character recognition (OCR) and Natural Language Processing (NLP) tools. The software can then be deployed with these capabilities to process structured or semi-structured paper-based content, such as invoices and extract information from emails and other documents. UiPath also claims to be in the process of adding machine learning capabilities that would allow customers to develop learning software robots.
Below is a short 2-minute video demonstrating how UiPath Enterprise RPA Platform works:
UiPath claims to have helped an unnamed global investment bank automate trade matching and settlement. Trade-matching is a key part of the trading process for investment managers, where trade details between the client and its broker are compared with each other. If matching is not performed at the right time, the trade may require human intervention or in some cases may fail.
The investment bank reportedly was facing high error rates and cost of operation in identifying unmatched or pending trades using their human analysts. Their manual trade matching processes involved a lengthy investigation process and email coordination between several business units. The bank deployed UiPath’s RPA solution to automate the trade matching process.
The Uipath software downloaded email from the system pertaining to trade executions and automatically matched the trade details to any pending or unmatched trades. The robot then updated the transaction numbers for any trade matches that had been identified in a spreadsheet. The robot also sent an email to the trade counter-parties.
According to UiPath, this resulted in the average handling time to process trades from 40 minutes using human analysts alone to 3 minutes after the integration of the software. We could find no information on what goes into calculating the AHT or what other initiatives were undertaken by the investment bank simultaneously.
We were unable to find any mention of enterprise-level companies on UiPath’s website nor in any of their press releases, but they have raised $408 million in funding and are backed by CapitalG, Sequoia Capital and Accel.
We found evidence of UiPath having machine learning engineers in its employee base, but we were unable to locate any C-level executive in the company with a history of AI experience.
Blue Prism is a London-based company with over 400 employees. The company offers an RPA platform which it claims can help financial institutions businesses automate business processes.
Blue Prism claims financial institutions can deploy their RPA platform to automate parts of their internal business processes and can use AI to create smart software robots that improve operational efficiency continuously. The Blue Prism RPA Platform can help finance companies monitor and execute scheduled tasks for their software robots. To this end, Blue Prism, announced a Technology Alliance Program in partnership with Appian, Captricity, Celaton, Expert System, IBM, and Minit. These companies are using Blue Prism’s Robotic Process Automation (RPA) platform to develop such AI RPA solutions.
Below is a short 5-minute video demonstrating how Blue Prism RPA Platform works:
Blue Prism claims to have helped The Co-operative Banking Group automate 10 internal business processes. The bank wanted to save costs by hiring fewer analysts and to speed up customer service operations. The bank identified 10 business processes: Direct Debit cancellation, account closures, CHAPS payments, foreign payments, audit reports, Internet applications, and Card and Pin Pulls.
These operations had a high level of human intervention and a large number of human workers managing the process. The banks deployed Blue Prism’s RPA software individually to each of these processes. According to Blue Prism, this resulted in the bank being able to ease the workload on their manual staff and lead to time and cost benefits in all the business processes within 12 months of the integration.
Blue Prism also lists Fidelity Investments, as some of their past clients.
We were unable to find any C-level executives with AI experience on the company’s team.
Takeaways for Business Leaders in Finance
Based on the companies discussed in this report, Robotic process automation seems to have traction among large financial institutions and banks where the scale of business processes is too large for human staff to deal with. In such cases, the financial firms can benefit by freeing up the time their staff spends on managing business processes and reduce operational costs.
UiPath has raised over $408 million in funding so far and has several established RPA use-cases for financial businesses. Blue Prism also lists several robust case studies on their website in addition to having raised over $59 million in funding. UiPath seems to be developing in-house AI enhancements for their software robots while Blue Prism seems to be following a partnership strategy, forming an Alliance with AI vendors such as IBM and Captricity.
Business leaders in finance who are familiar with RPA might be interested in adding artificial intelligence capabilities to their software robots to optimize their functioning over time. In the next two to five years we expect that business processes involving document digitization (extracting information from non-digital formats like paper) might be automated using intelligent OCR and NLP. We might see RPA platforms leverage machine learning to automatically prompt businesses with insights on improving efficiency. RPA might still be a necessity only for firms with large enough scale of operations where the integration and capital costs are justified by the cost savings achieved through the automation.
Header Image Credit: Dan-Bul Study