Building a Fully Serverless MLOps Platform at Aviva with AWS Enterprise MLOps Framework

In today’s digital age, the integration of machine learning (ML) into various industries has revolutionized how businesses operate and manage their data. One such industry that has embraced ML technology is the insurance sector, with Aviva leading the way in implementing innovative ML solutions to enhance their operations and customer experience.

With a rich history dating back to 1696, Aviva has evolved into a global insurance powerhouse, serving millions of customers across 16 countries. To stay ahead of the curve and meet the growing demands of their customers, Aviva has turned to ML to streamline their processes, improve efficiency, and provide personalized services.

In collaboration with AWS, Aviva has successfully built a fully serverless MLOps platform that leverages the power of AWS Enterprise MLOps Framework and Amazon SageMaker. This platform enables Aviva to standardize model development, streamline deployment, and ensure consistent monitoring of ML models across their organization.

The challenges faced by Aviva in deploying and operating ML models at scale are not unique. According to Gartner, nearly half of all ML projects never make it to production, highlighting the need for organizations to establish robust processes, effective monitoring, and invest in the necessary technical foundations for successful MLOps implementation.

To address these challenges, Aviva chose their Remedy use case as their first project on the MLOps platform. This use case involves a data-driven approach to managing car insurance claims, using 14 ML models and business rules to determine whether claims qualify as total losses or repair cases. By successfully deploying and evaluating the Remedy use case, Aviva has set the stage for future ML use cases, demonstrating maximum efficiency and scalability.

The MLOps platform implemented by Aviva is built on a foundation of best practices, including reusable ML pipelines, serverless workflows for orchestrating model inference, and robust monitoring capabilities. With a focus on security, Aviva ensures the protection of customer data and intellectual property, using encryption, access controls, and auditing mechanisms to safeguard sensitive information.

The partnership between Aviva and AWS has resulted in a scalable MLOps platform that accelerates Aviva’s ML journey and reduces infrastructure costs by 90%. By harnessing the power of AWS Enterprise MLOps Framework and Amazon SageMaker, Aviva has positioned itself as a leader in using AI technology to drive innovation and efficiency in the insurance sector.

To learn more about the AWS Enterprise MLOps Framework and how it can accelerate your organization’s MLOps journey, we encourage you to explore the platform on GitHub and discover the endless possibilities of integrating MLOps into your organization.