Aviva Migrates its Petabyte-Scale Data Estate to Snowflake to Deliver Even Better Customer Experiences

With over 25 million customers and a history that spans more than 325 years, Aviva is a leading diversified insurer in the UK, Ireland and Canada, and data is core to its business operations. 

Aviva has always maintained a laser focus on data-driven customer journeys, underwriting, managing claims and decision-making. However, as the organisation grew and acquired other companies, its data teams faced challenges with managing and maintaining several different data platforms. 

As it looks to make the most of new technologies, like generative AI and near-real-time data streaming, Aviva embarked on a radical transformation programme to shift its petabyte-scale data ecosystem to Snowflake’s AI Data Cloud.

When Aviva began its data platform modernisation journey, it established strict criteria for any potential vendors: It needed a cloud-native stack, support for multiple cloud environments, componentised architecture and flexible connections that would avoid vendor lock-in.

“Snowflake met all these key deliverables,” explains Niall Scott, Head of Data Engineering and Data Operations at Aviva. “But it went beyond that, with a clear roadmap for delivering regular updates that would keep us on the cutting edge of what’s possible with data.”

Aviva started its Snowflake journey with the launch of Aviva Zero, an agile car insurance product, which runs entirely on Snowflake's AI Data Cloud and helps policyholders offset their carbon emissions from driving.

Following the success of Aviva Zero, the insurer began shifting core workloads from Oracle Cloud environments to Snowflake on AWS, starting with key claims and fraud detection use cases. As an organisation with a sizeable data estate, this required Aviva to migrate petabytes of data over several months — without impacting the high quality of service its customers have come to expect.

By partnering with Snowflake’s Professional Services team and AI Data Cloud Services Partner NextPathway, Aviva had upgrade partners to help design and implement its Snowflake estate and operating model. This ensured the company was in the best position to hit the ground running and scale its infrastructure as it migrated more data workloads to Snowflake.

“We were able to lean on the Snowflake Professional Services team extensively thanks to their deep understanding of the insurance industry,” says Niall. “They also brought expertise from outside our domain as well. We see a lot of value in being able to see how other industries are operating and adopting their best practices to ensure we can scale effectively.”

With support from Snowflake’s Professional Services team, Aviva migrated a large portion of its data estate, totalling 1.3 PB, to Snowflake on AWS in just seven months. Once this key workload was moved, Aviva’s data team quickly saw an eight times improvement in data processing speeds. “We can physically process data much faster,” explains Niall. “Because everything is in Snowflake, there’s no requirement to transfer data externally. We used to run microbatches every five to 10 minutes, but now we can stream data to see what’s happening in near real time. That’s been vital in supporting our customers, particularly when they’re most vulnerable.”

With Snowflake’s per-second billing structure, Aviva also gets faster processing at a lower cost, as compared to its legacy data estate. “With Snowflake, cost is based on usage, which has led to a reduction in costs,” explains Niall. 

Since moving to Snowflake, the Aviva team has saved time and helped reduce the management burden on its engineers, as the AI Data Cloud is entirely updated, patched and managed by Snowflake. “Snowflake is a technology company, and we’re not,” says Niall. “They’re best placed to support the technology and keep us current with the latest features and tools. With things like up-to-date Python libraries, we always have the tools we need to deliver great customer experience."

The speed of Snowflake has been beneficial for Aviva’s decision-makers and data analysts, but the platform can also help to improve its service to customers. For example, when repairing a customer’s vehicle after an accident, Aviva teams can interrogate data quickly to find a nearby garage that will effectively repair the customer’s vehicle and avoid additional mileage.

By bringing multiple workloads into Snowflake, Aviva also gets a more complete view of customers across insurance lines, pensions and wealth management teams. This means the insurer can understand customers’ needs faster through using complex multi-product modelling to provide a more connected service experience.

Aviva already has big plans for the future with Snowflake. Its first focus is on new innovations like generative AI. Aviva has already deployed gen AI to quickly summarize complex claim journeys for contact centre agents, but the insurer wants to take this further and find even more use cases.

Aviva intends to become even more flexible with the existing Snowflake tools it relies on to keep delivering the smoothest customer journeys. “And we’re using the data we now have in Snowflake to drive our organization forward. We’ve always had an ambition to be a truly data-driven business. Now, with Snowflake, we have the tools and data we need to deliver on that ambition.” says Niall.