Key challenges in RPA and AI development

Saravanan Kumar, Aon’s global head of automation, gives his take on the space

Although the adoption of robotic process automation (RPA) and artificial intelligence (AI) is in its infancy in this region, the emphasis on these emerging technologies at this year’s GITEX event is a clear indicator of their importance and the need for businesses to engage with these technologies. Saravanan Kumar, the global head of automation at Aon, will be offering his real-world insight into these fields at this year’s conference and sharing his expertise on the framework he has developed. His hope is that CIOs can use this framework to build capability and credibility in order to bridge the gaps between technology, business and operations to ensure business buy-in.

Kumar explains: “The model that we use for our automation platform is envision, engage and enable, [which leads to] excellent.” The progress through these stages involves working with the business as a whole to take into account its key priorities and challenges, to identify the best technology available and then to define the capabilities that need to be built. The framework covers a two-year period from inception to the final stage excellence.

According to Kumar, many businesses are concerned about a bottleneck in developing these technologies because of security issues, a lack of tech availability, or feel daunted by potential challenges. Yet he is clear that by engaging with all aspects of the business, the technology can be introduced successfully. Recalling a discussion with a delegate at a London RPA conference, Kumar explained that he had been asked the best way to successfully implement bots in an organisation. He replied: “‘If you really want to be successful with bots, you need to start with bots.’ I explained that if you want to make a bots implementation successful you need business, operations and technology (BOTs) to come together.”

Kumar believes that simply creating a Centre of Excellence (COE) does not provide the solution though, as a change of personnel can lead to wasted time and effort. His solution? To create a centralised, intuitive and capable RPA platform and systemise it. Aon is trying to create just such a platform, centralising capabilities from finance to front office and everything in between, so that users can turn to it as the starting point for their own innovation. In describing the aims of the Aon platform, Kumar draws a comparison with the Uber business model: “We want to ‘Uber-ise’ the RPA platform Uber doesn’t have a fleet of cars. It has a technology platform that brings together the person with the car and the person who needs a car.”

Explaining the benefits of the RPA system in the context of Aon, Kumar explains that applying AI and machine learning to his IT operation reduced their work by 85%. He says: “We have a large footprint of systems across the globe ¬ we’re in more than 150 countries. We have servers and networks across the globe…we get more than 130,000 alerts from these systems. So it is literally impossible to put a person [there] to manage these events. To solve the problem of fixing the network, instead of calling different people for storage services…we have made algorithms that can learn [and] understand, even if the servers, the databases and storage associated with that network fail. So the system should be intelligent enough to understand a pattern of events and then give an intelligent suggestion; saying what it thinks the next link [step] should be. We started working on these concepts and we were able to show a significant reduction in events.”

The RPA project at Aon is expected to last three years, as Kumar advises simplifying and optimising programmes before automation. “I am a strong believer that if you automate garbage, you only automate inefficiencies. Unless you clean it up and do the right thing you are not going to get the desired results. This is a great opportunity for the company to decide [whether] this is the right process, why [it is] doing this and build some next gen, digital platforms. The analogy I tell my team is that every problem you see is a boulder. … A big boulder needs a sledgehammer.”

Saravanan Kumar will be appearing in the banking stream of the GITEX Conference