RPA - Task automation with high productivity

Since the advent of computer, many kind of application software - for instance, MS Excel - helping enterprise employees to perform their tasks smartly in digital way. However, employees are still performing many tasks using digital systems that are repetitive, monotonous and high volume in nature. Many of these tasks are well structured and have clearly defined business rules. Doing these tasks with human workers are error prone and usually take very long time. With high error probability, it may cause heavy penalty to the enterprise, if this error breaks any regulation of the control body.

Recently, along with the wave of 4IR technology such as AI, Block chain, VR, AR and more - has emerged a new technology, the software robots that can work as complementary digital work force of human workers in the office. These robots can perform the repetitive, high-volume and structured tasks accurately and efficiently with fraction of the time required by human workers. This technology is called Robotic Process Automation - in short, RPA. RPA can be complemented with AI technology such as Computer Vision (CV) and NLP (Natural Language Processing). As a result, this combination of technology makes possible to automate more complex tasks in the business process work flow.

Consider a regulatory service provider company that provides monthly invoice processing service to its enterprise clients. The way it implements this service is shown in the following business process flow:

A back office operator checks if the customer submits the invoice data (for instance, an Excel file). If delayed, operator sends reminder to client. After receiving invoice file, operator validates and import the invoice data. In case, validation error, operator sends the error log to the client by email. Otherwise, operator generates the report from system and provide it to the client.

Most of the activities in the process can be automated easily with RPA. Consider the updated business process where back office operator and RPA robot work together to complete the invoice processing.

In this updated work flow, a RPA robot checks if the customer provides invoices files within the predefined schedule. If delayed, it will notify the customer. After receiving the file, RPA robot will validate data with defined rules and in case of error, it will send the error log to the client. Otherwise, it will import invoice data to the system and generate the report. Backoffice operator will review the report. When the review status is okay, RPA will send the report to the client.

If this regulatory service provider has thousands of clients, you can easily imagine how much human effort can be saved with the RPA robot. Moreover, it can process invoices 24/7 with 100% accuracy.

This automation can easily be extended with the help of AI CV (Computer Vision) and NLP (Natural Language Processing) technology. In the preceding example, many customers will provide invoices with their own company format in the PDF file. AI CV technology can be used to learn these various invoice format and extract data from the invoice. Using NLP, system can understand the invoice content. Hence, RPA robot equipped with these new capabilities can process data from different invoice format automatically.