Robotic Process Automation and Artificial Intelligence

Posted on: 01 November 2017

Download this article [PDF]

In order to respond effectively to ever increasing customer demands, companies are investing in technologies which can allow products to be brought to market in a more agile way. This is where the introduction of automation comes into its own.

Due to the number of manual processes found within a typical financial company, it is not surprising to hear that this sector is one of the leaders in the use of automation technologies.

Rather than being something for employees to fear, many experts are quick to reassure that automation will instead have a positive influence for our workforce, with many of us interacting with automation on a daily basis along with our colleagues.

Robotics Process Automation (RPA), one of the first steps in the automation journey, is already showing benefits across a variety of industries. For example, banks and financial institutions are deploying RPA to optimise their back office processes; the healthcare profession is exploring the use of RPA for patient registration, and human resources departments have utilised RPA for compliance reporting.

Automation using RPA is typically being used for back office tasks which follow a strict process and do not require any understanding of the process being automated. This type of activity is commonly seen by humans as boring and repetitive. The process of automating this type of process is often referred to as 'taking the robot out of the human'.

It’s important to remember that RPA is not suitable for all processes, and sometimes only part of a process can be automated. While the processes need to be clearly defined without the requirement for cognitive decisions, this does not mean the robot cannot carry out logical rules such as mathematics, comparisons and pattern matching. The limitation of RPA is that it does not understand the activity it carries out and is simply following a predefined list of steps.

Harder, better, faster, stronger

So what are the advantages of RPA? According to Accenture, an RPA ‘robot’ acts as a super-efficient human in a number of ways. Tasks suitable for automation can be performed in a fraction of the time, at a fraction of the cost with claims of a time reduction of up to 90% on tasks. The company Blue Prism, which specialises in RPA software, cites The Co-operative Banking Group amongst its clients benefitting from RPA through automation of their payment queuing process, ultimately improving efficiency by 80%1. This activity was previously undertaken by nine employees, reviewing 2,500 customer accounts.

RPA also eliminates the scope for human error. Robots don’t become distracted, weary or ill and they can work 24 hours per day without the need for breaks. As a result, automated processes are fully auditable and provide a higher degree of accuracy and compliance to internal frameworks.

Process automation implementations can be measured in weeks with no requirement to make changes to applications. This is because RPA runs on existing IT infrastructure and works at the graphical user interface level, needing little or no IT support. Instead the main considerations involve identifying appropriate processes to automate and implementing them with effective change management and communications.

Impact on workforce

So where does this leave employees? Earlier this year a large insurance company received press coverage for allegedly asking its workforce if they could be replaced by robots, no doubt adding to fears that automation means redundancy.

However, far from facing elimination, the trend is that employees will be freed up from tedious, uninspiring tasks in order to work on more value-add, skilled activities.

The professional services firm EY claims that through engaging employees in more rewarding activities, attrition will be lower and productivity will be increased. In the previously mentioned Co-operative case study, employees were re-deployed into customer account management roles, which in turn drove up outbound customer service call levels. Also, Deutsche Bank makes reference to augmenting its workforce through RPA, citing the reduction of mundane tasks resulting in lower attrition. The message coming through is clear; this advancement is about supporting and enabling human employees rather than swapping man for machine.

RPA and beyond…

The next step in the automation journey is IPA (Intelligent Process Automation). RPA is designed merely to follow instructions, whereas IPA which is made up of RPA and AI (Artificial Intelligence) has the capability of learning from experience which it then applies to improve on future tasks.

Almost without people noticing, AI is fast becoming part of our daily life. Examples include the mobile phone ‘personal assistants’ Siri and Cortana; Amazon’s 'Alexa' voice service, and several types of fraud detection software used by banks.

Currently AI remains fairly immature, but evidence suggests that this disruptive technology will transform many industries, including insurance. A number of start-ups are capitalising on this prediction, for example, the New York-based company Lemonade, which sells insurance using a chatbot smartphone app and exploits automated algorithms to great success for speeding up its claims service.

This is one example of AI but for insurers, the possibilities seem endless. From spotting patterns and suspicious activity for fraud detection to using virtual assistants with the addition of chatbots for a more natural interaction experience and finally self-service robotics, which provides the capabilities for websites and mobile applications to be dynamically created, and allow RPA/IPA to serve customer requests directly.

For an industry typically viewed as staid and old-fashioned, this may present an opportunity for many insurance companies to rebrand themselves as leaders in digital innovation.

Robots and brokers

How is the role of the broker likely to change alongside these advances in technology? While the simpler processes may be automated, the prediction from companies such as Konsileo, a start-up commercial insurance broker, is that larger SMEs and corporates will continue to rely on human brokers for their advice and recommendations. This is mainly due to the complexity of the potential risk involved.

Importantly, the broker role also focusses on flexibility and communication, meaning that many customers still prefer human methods over self-service online processes. However, some change in working styles will be needed and brokers should consider how RPA and AI can be used to enhance their value proposition.

How could RPA and AI assist insurers and brokers?

  • Fraud detection
  • Predicting market insight and customer behaviour
  • Social network analysis
  • Risk assessment and selection
  • Categorisation of damage
  • Virtual personal assistances/chatbots
  • Underwriting
  • Claims
  • Data collection
  • Help with pricing

Through intelligent understanding and application of Big Data, brokers can gain ever more insight into consumer needs and behaviour. AI, especially when augmented with Big Data, could be the tool to equip brokers with more information and knowledge to better serve their customers.

Download this article [PDF]

You may also be interested in some of our other features related to automation and other technological advances that could affect everyday lives, such as:


1 Blue Prism/Co-op case study