As Australian businesses rethink their operations in the wake of the pandemic, the big question is not “Where will people work?” but “How can we adapt fast enough to keep up with our new reality?”

Rolling lockdowns and ongoing uncertainty mean the dynamic nature of the business environment has already put great strain on traditional processes and systems. As the second and third waves of the pandemic hit, we can expect more surges and drops in demand, supply chain disruptions, and ongoing shifts in both customer touchpoints and workforce locations.

In this environment, we know that human response times are simply not fast enough. When the pandemic hit, the companies that spun up fastest were those using AI to augment their human teams.

In India, Accenture moved 150,000 people to work from home – in three days. In that time, every single person was contacted about what they did and didn’t have and technology was procured and delivered to their homes. In the weeks that followed 150,000 people were checked on personally every day and given trouble shooting advice to improve network performance.

The only thing that made this possible was ‘AI-infused service delivery’ – where bots did the heavy lifting, performing millions of communication and coordination tasks, and the humans in the process were fed information by AI.

Otherwise, it would have taken those three days to mobilise the extra call centre operators required to resource this vast project – and who knows how long before every person was working effectively at home.

AI-infused service delivery is a relatively new approach based on patented technologies that: auto-convert knowledge to digitised steps; proactively identify issues, self-heal and learn; and resolve issues either by generating their own code or drawing from an automation library.

The difference made is this. A world-class human team trying to shift organisational responses at scale takes three to five days before the smallest change occurs. With AI-infused service delivery, that same change takes three to five minutes.

Companies around the world are using this approach to make 50% reductions in operating costs, 70% faster tickets resolutions and 80% improvement in SLA responses.  

Its power comes from solving the ‘dilemma of smart things’. With so many different devices feeding disparate parts of data, and so many services in the cloud, how can companies coordinate and synthesise all this information to find the right actions? AI-infused service delivery answers this question by building smarts into that ‘data fabric’, organising it so AI (and humans) can see where and why things are failing – and move quickly to correct them.

For example, a US retailer wanted to know why some of its 250,000 point of sale (POS) devices were periodically failing. With 10,000+ stores, each with 50+ different wireless access points, the data coming off the problem was a vast wall of ‘noise’. But building an intelligent layer allowed us to discover why POS devices were failing, automatically rectify the issues – or use a virtual agent to prompt retail staff to fix the problems on site – and, eventually, to predict and prevent failures before they happened.  

When we first started investing in automation, people were directing machines. Now, machines are smart enough to support us. Rather than a human triggering a robot, the data can drive AI to drive people to confirm an action, provide missing information or make a decision.

The big advantage here is the human is only told what they need to know. We don’t get swamped by the overwhelming volumes of data required to surface the insight.  All we have to do is to use our human smarts to respond to it.

It’s an evolutionary step in our relationship with AI – and the only way businesses will be able to respond to the wild swings in demand that are now business as usual as we go back to work in a post-COVID world.

Luke Higgins

Managing Director, Growth Markets Automation Lead

No posts available at this time.
Subscribe to Accenture's Anztrends Blog Subscribe to Accenture's Anztrends Blog