Suncorp is exiting data centres by 2023

Suncorp is exiting data centres by 2023

This final exit will be an historic shift for both the insurance giant and the industry. For decades, data centres have housed insurance companies’ computer and storage systems. Since 2013, said Pizzato, his firm has pursued an ambitious cloud strategy that has involved transitioning away from data centres.

Moving to the cloud

“We got on the stage of AWS re:Invent Vegas and set a pretty ambitious target of moving a number of workloads to AWS over a six month period,” he said.

AWS, or Amazon Web Services, describes its annual re:Invent Las Vegas gatherings in the casino world capital as a “learning conference for the global cloud computing community.”

In the years following that announcement, said Pizzato, the company moved 65% of its computing, including programs and applications, to cloud environments.

“So that’s a broadly public cloud in AWS but also a number of services in Microsoft Azure and we use software as a service with partners like Google,” he said.

Read more: Are insurers too dependent on global cloud technology firms?

Pizzato said “a significant amount of our critical assets” are in Google Cloud.

“The reason for that back then was largely to get infrastructure off the critical path of the technology lifecycle,” he said. “So, to ultimately be able to speed up the delivery of service to customers.”

Pizzato said it has also helped with the deployment of software, for example, having the capacity to scale services up and down rapidly as needed.

“You can build without the sort of constraints of the traditional data centre and the lead times around the various components of it,” he said.

See also  What's going on with builder's risk?

Changing challenges and opportunities in the cloud

However, in recent years, financial services, including the insurance sector, have grappled with new digital and data opportunities and challenges that have changed their use of the cloud.

“The problem has changed from initially being, how do we speed up software? To opportunities around how we use data for innovation and effectively leverage services from those cloud providers,” said Pizzato.

“We’ve now moved from having 65% of our workloads to 90% of our workloads in the cloud,” he said.

Read more: Multi-cloud – which one should insurers choose?

The remaining 10%, said Pizzato, are services that can’t be put into a “hyper scaler” like software and hardware systems from Oracle Engineered Systems, the American multinational technology firm. Those will remain in a facility in Sydney.

One example of where the increasing use of cloud-based technologies with AI (artificial intelligence) and machine learning tools is helping Suncorp, said Pizzato, is around geospatial technology.

“We use that in our Disaster Response Events Centre when we have claims events,” he said. “What that lets us do is overlay a weather event on a map of all of our risk within a certain area and effectively show, as a weather event is passing through, where our at-risk customers are.”

Pizzato said this visualisation, on a wall size screen in their Brisbane Disaster Recovery Centre, shows the path of the event, the risks, where houses are located and the damage.

“You can never consider doing that if you weren’t working with partners that are leveraging the scale of the cloud,” he said.

See also  HDI Global SE reveals financial results for H1 2023

During and after a claims event, said Pizzato, this technology allows disaster response teams to proactively engage with impacted customers.

“After the event, it allows us to understand very quickly what the damage has done to certain properties and that speeds up the claims process significantly,” he said.

Read more: Is insurance industry data really safe in the cloud?

Pizzato said geospatial technology also allows claims teams to contact a customer impacted by an event and suggest what the claims outcome should be – before they’ve lodged a claim.

According to Rackspace cloud computing expert Angela Bartels, data centres have their roots in the huge computer rooms of the 1940s that were largely used by government agencies. The boomtime, said Bartels, in her article “Data Center Evolution: 1960 to 2000”, came during the dot.com bubble of 1997-2000.