Digitising expertise

Digitising expertise

Cognitive automation promises a great deal… but how do you get started, and exactly how can a process be transferred from human workers to software? Oliver Franklin-Wallis interviews Vinodh Swaminathan, Managing Director of Innovation and Enterprise Solutions at KPMG in the US, to find out.

Managing Director, Innovation & Enterprise Solutions

KPMG in the U.S.


Also on KPMG.com

digitising expertise
  • New AI offers firms much higher level of 'human' judgement
  • Cognitive automation shortens the time it takes for new hires to get up to speed
  • Many companies investing in AI but still in experimentation stages

As artificial intelligence matures, we’re entering an era in which a significant proportion of work hitherto considered inherently ‘human’ can be automated. 

But incorporating these new technologies into your business requires a major transformation from top to bottom. “This is not a technology discussion. This is a business strategy discussion,” says Vinodh Swaminathan, managing director of Innovation and Enterprise Solutions at KPMG. “It starts at the top with leadership and stakeholder management.”

Amplifying human expertise
The first practical step is to identify the areas within a business where automation will offer an advantage. “You need a sense of what your people do, where they’re spending a lot of their time, and which of those tasks qualify to be automated,” says Swaminathan. “You can then pick off tasks and start to automate them.” 

The easiest tasks to automate are routine, repetitive and well-codified. In the past, enterprises have often found that these tasks are not as straightforward as originally envisioned, therefore still requiring significant human oversight. However, the cognitive technologies becoming available today solve this problem by bringing a much higher level of ‘human’ judgement and expertise to task automation. 

Similarly, enterprises can identify areas where existing expertise could be amplified and augmented with the help of cognitive automation. 

One example might be proficiency training: using AI systems to augment the capabilities of new hires – for example, in call centres. “With these machines you can significantly shorten the time it takes for a new hire to reach proficiency,” says Swaminathan. In other areas, cognitive automation can diagnose problems and suggest smart solutions – such as processing back-office payments, parsing large volumes of legal research or narrowing down medical diagnoses. 

A new kind of IT project
As with any such fundamental shift, it’s important to identify the individuals and departments who will be managing the transition. “This is not a traditional IT project,” says Swaminathan. “The chief information officer will always play an important role because these are all technology deployments and use enterprise data. But you don’t install these in the way that you would install your e-mail – given the enormous input from domain experts in training these systems to mimic human thinking, the business typically ends up playing a more significant role in deployment.”

Be prepared: incorporating these new technologies into a business is not an overnight process. “As we speak, IBM is trying to recreate a physician workflow in an oncology environment. That’s an 18- to 24-month process – and even that doesn’t get you all the way,” says Swaminathan. “But if you’re looking at creating a paralegal or a customer-service representative, that’s relatively straightforward compared to getting an oncology exam.” 

Taking the experimental approach
While some large companies are already bringing products to market, such as IBM with its Watson cognitive system, it’s also vital to stress that technologies like deep learning and AI systems are new and emerging – and bring challenges with them as a result. “There is a degree of learning and experimentation,” says Swaminathan. But as industries inevitably begin to adopt digital labour, the rewards for success will be vast. 

“We have plenty of recent history: you look at the web, you look at e-commerce, you look at mobile, you look at social. We know that you cannot sit on the fence and wait for these things to play out,” Swaminathan says. “We don’t believe ‘fast follower’ is the right strategy for cognitive automation. Cognitive automation is fundamentally about how knowledge capital is deployed in a disruptive way – playing catch up once your industry has been disrupted is much harder. You really want to be the innovator in the early part of the cycle, because the investment will pay off in the form of a very significant competitive differentiation.” 

The most successful businesses in the future will be those that embrace new technologies. “You’ve got to take an experimentation approach – companies that fail to prepare now risk being left behind.” 

Connect with us


Request for proposal



KPMG’s new-look website

KPMG’s new-look website