Kirill Kedrinski - Fotolia
Many activities in an organization's finance department are manual and time-consuming, including executing processes, extracting data, building reports and solving problems arising from the manual processes.
Today, robotic process automation (RPA) is changing some of these activities by enabling enterprises to automate repetitive and manual tasks. In doing so, RPA is helping to streamline and potentially transform the finance department.
RPA in finance
RPA works well for simple processes that operate in relatively high transaction volumes -- and finance and accounting are ripe with them, said Craig Le Clair, vice president and principal analyst at Forrester Research.
"One bank that I interviewed had 1,400 people closing the books monthly, quarterly, end of year, and they felt they could automate [the work of] about a third of those full-time employees with RPA."
Imran Sabir, the senior manager of RPA at OZ, a consulting company based in Fort Lauderdale, Fla., agreed that RPA can improve an organization's end-of-year closing, which is the most hectic time for finance.
The financial close and reporting process encompasses numerous tasks that involve many systems, departments and individuals, from closing out subledgers to creating and delivering financial filings to regulatory bodies, Sabir said. The process requires posting data from sources such as Microsoft Excel to these subledgers -- a tedious undertaking that RPA can mitigate and solve efficiently.
Reporting is another common use case for RPA in finance, according to Sabir.
Reporting manually involves data gathering, filtering, extraction and cleansing. In addition, data residing in other systems has to be integrated with the finance data. Once these reports are automated with bots companies can save a lot of time and effort, Sabir said.
For many companies, adding a new supplier or account can also be a manual process. By implementing RPA in finance, a bot can do this in an automated way, which can help companies save time, he said. For instance, a bot can get information about a new supplier automatically -- e.g., credit scores and tax information. Additionally, bots can be used to automate the process of vetting new customers and validating and updating their data.
The starting point for RPA is structured data, according to Chris Huff, chief strategy officer at Kofax Inc., a process automation software provider based in Irvine, Calif. However, the majority of data in the finance department or CFO function is in an unstructured format -- i.e., sentences and paragraphs in a document or words in an email.
"How do you transform that unstructured data into a structured format so that an RPA finance digital worker can then execute the automated task? What we found out is that we can add what Kofax calls in our intelligent automation platform 'cognitive capture' to lift all this unstructured data and put it in a structured format and then give it to the RPA robot," Huff said.
Extracting unstructured as well as structured data from various forms from source systems such as other websites and placing that data into target systems like accounts payable and accounts receivable is one of the functions of RPA, said Carl Lehmann, principal analyst at 451 Research.
RPA does this by automating document processing using scripts and machine learning technology, typically natural language processing and optical character recognition, he said.
"There's a lot of machinations that need to occur when a finance department has multiple systems," Lehmann said. "And that leads us to the second thing that RPA does: use machine learning technology to examine what workers do on their screens."
RPA benefits the CFO
Part of RPA is called process mining and part of process mining is device mining, Lehmann said. Using RPA in finance and device-mining technology, a CFO can examine the screens, forms, applications, websites -- everything workers use -- to understand how they do their work, what works well and what doesn't.
"That technology exposes a visual pattern of the various tasks that a worker performs," he said. When those tasks are recognized as patterns, they can be analyzed for efficacy or inefficacy. Then, once they're analyzed for efficacy, they can be automated using a software bot. And that leads to the third thing that RPA does.
All of a sudden, the software bots that perform the basics become a collection of bots that operate collectively and become a digital skill -- the same skills a human has, according to Lehmann. The skills could be in the auditing process, for example, or loan origination or strategic sourcing. The bots could assemble information about a specific supplier to determine whether it qualifies as a preferred supplier or meets certain criteria for advance payment, he said.
These three fundamental things about RPA in finance enable the CFO and the CFO's organization to do the tasks they need to do, such as rolling up the books for reporting to expose the relevant data needed for regulatory compliance, he said.
Bots can also help the finance department with currency exchange, said Chris Koeneman, vice president of strategy at Tangoe Inc., a provider of cloud expense management software based in Parsippany, N.J.
"It used to be that you'd have to have some sort of currency exchange engine running inside of your software to figure that out," he said. "Well, there's a treasure trove of information on currency exchange rates, which are very complicated. But you can get that information for free from a plethora of sources with a bot on the front end of your financial system. And you wouldn't have to pay a service bureau to do it."
The biggest advantages of implementing bots are in the complex business areas, such as finance, supply chains and operations and aftermarket support, OZ's Sabir said.
"These business areas are usually well-documented with most of the work being done manually," he said. "Automating these processes can save thousands of hours of manual work, and companies can get things done much more quickly. In the end, leading to more positive experiences internally and externally."