A Case Study


Automating Processes with Robotic Process Automation

Process automation is a hot topic across enterprises these days. 68% of CEOs have asked their operations and IT departments to leverage robotic process automation (RPA) to automate low skilled labor, while looking to machine learning (ML) and/or artificial intelligence (AI) to automate medium and higher skilled labor (footnote 1). In fact, by 2020, over $3B will be spent on automating processes at companies of all sizes (footnote 2).

 Automating a process begins with an assessment of the process complexity. The two key factors to consider when deciding what automation path to take are: (1) the availability of business rules that are easy to discern and apply; and (2) the availability of skills and knowledge required to complete the process. Where the business rules are obvious or well documented and where the skills required are basic, RPA is often the best tool for process automation. When a company must develop and define a more complex set of rules or the people participating in the process require significant on-the-job experience or higher education, leveraging ML and AI are the way forward.

 For this case study we will focus on RPA, through which most business processes can be automated.

 “This all sounds great, but give me a real-life example, please”   

The Accounts Payable (AP) process is often one of the first processes that companies automate. This was true at a large multinational insurance carrier which did not have a history of leveraging robotics at scale prior to this experience.  The insurance carrier uses SAP for finance and accounting along with some home-grown software solutions. The insurance carrier found that processing supplier invoices through the system required several manual interventions: reading invoices, validating math and vendor information, and registering and posting invoices across multiple systems. As part of an efficiency drive, the insurance carrier decided to use automation within its Finance AP team, which processed over a thousand vendor invoices every day.

By implementing RPA, the insurance carrier managed to eliminate almost all human intervention from the AP process. As RPA software mimics human actions, it operates on top of existing computer applications and interacts with these applications in the same way a human would via the native user interface. This means no changes to applications and a drastically reduced implementation time.

At the insurance carrier, the AP processing robot logs into the necessary systems, reads the invoice image, registers the invoice in SAP, performs a set of validation steps (including cross-checking against other systems), and ultimately decides whether to post the invoice. Having a single robotic solution follow the end-to-end process flow and communicating with all the systems involved, brings much higher automation benefits than the traditional mix of humans and partial automation.

The insurance carrier has gained significant benefits through RPA. The insurance carrier’s personnel are spending 65-75% less time on manual tasks and the data quality and error rates have been reduced as well.  An unexpected benefit was the digitization of the AP process which has meant structured access to data about the process which is used to continuously improve and further automate the process.

This means that RPA has not only directly improved outcomes and the cost-base, but it has also ensured further improvements thanks to process insights and performance transparency.

The AP Process Pre & Post RPA

The AP Process Pre & Post RPA


“Were there any lessons learned?”

People may perceive robotics skeptically, since it is often viewed as an expense play at companies. As a result, it is important to be very focused on communication throughout an organization, being inclusive as RPA is developed and deployed, and being focused on change management before, during, and after an implementation. At the insurance carrier it was important to communicate that the robots were not taking away jobs. Instead the robots were used to allow people to focus their time on value-accretive activities where they previously were not able to spend substantial time. Additionally, employees were heavily involved in testing the automation solutions to ensure that the output of the robots were fulfilling the insurance carrier’s needs.

Every organization has made attempts at automating process. However, most attempts have resulted in mixed results or tactical solutions (e.g. excel macros, MS Access databases) that fail to meet the standards a corporation’s modern control environment requires. As a result of leveraging RPA solutions, companies can adhere to regulatory and audit requirements, constantly optimize their processes, and rapidly scale as robots can be turned on and off at the click of a button while never requiring training or retraining.




1.        Vivian Couto & Karina Fernandez-Stark, Duke University, “Jamaica in the Offshore Services Global Value Chain”, April 2018, https://gvcc.duke.edu/wp-content/uploads/2018.04.26_Offshore-Services-GVC_Jamaica_-public.pdf

2.        Deloitte Touche Tohmatsu Limited., “The Robots Are Ready, Are You?”, 2018, https://www2.deloitte.com/content/dam/Deloitte/us/Documents/process-and-operations/us-cons-global-rpa-survey.pdf