Accelerate Time to Business Value Using Amazon SageMaker at Scale with NatWest Group

NatWest Group is one of the largest banks in the United Kingdom. Formally established in 1968, the company has origins dating back to 1727. NatWest Group seeks to use its rich legacy data to innovate and personalize its personal, business, and corporate banking and insurance services. To deliver these solutions at a faster pace, the bank needed a standardized ML approach. “We didn’t have a consistent way to access our data, generate insights, or build solutions,” says Andy McMahon, head of MLOps for data innovation for NatWest Group. “Our customers felt these challenges because it took a much longer time to derive value than we wanted.”

To deploy personalized solutions at an enterprise scale, NatWest Group chose to adopt Amazon SageMaker as its core ML technology. The bank also engaged AWS Professional Services, a global team of experts that can help companies realize their desired business outcomes when using AWS, to prepare for the project. During a series of workshops, NatWest Group and AWS Professional Services worked together to identify areas of improvement within the company’s ML landscape and created a strategy for development. After crafting a comprehensive plan, the teams began working on the project in July 2021.
 

In April 2022, NatWest Group launched an enterprise-wide, centralized ML workflow, which it powers by using Amazon SageMaker. And because the bank already had a presence on Amazon Simple Storage Service (Amazon S3)—an object storage service offering industry-leading scalability, data availability, security, and performance—this was the service of choice for its data lake migration. With simpler access to data and powerful ML tools, its data science teams have built over 30 ML use cases on Amazon SageMaker in the first 4 months after launch. These use cases include a solution that tailors marketing campaigns to specific customer segments and an application that automates simple fraud detection tasks so that investigators can focus on difficult, higher-value cases.

NatWest Group employees now have fast and simple access to the data and tools that they need to build and train ML models. “We modernized our technology stack, simplified data access, and standardized our governance and operational procedures in a way that maintains the right risk behaviors,” says McMahon. “Using Amazon SageMaker, we can go from an idea on a whiteboard to a working ML solution in production in a few months versus 1 year or more.” NatWest Group launched its first offerings in November 2022, reducing its time to value from 12–18 months to only 7.

To accelerate its employees’ workflows, NatWest Group uses AWS Service Catalog, which organizations use to create, organize, and govern infrastructure-as-code templates. Before the bank adopted this solution, data scientists or engineers would need to contact a centralized team if they wanted to provision an ML environment. Previously, it would take 2–4 weeks before the infrastructure was ready to use. Now, NatWest Group can launch a template from AWS Service Catalog and spin up an ML environment in just a few hours. Its data teams can begin working on projects much sooner and have more time to focus on building powerful ML models. This self-service environment not only empowers data science teams to derive business value faster, but it also encourages consistency. “As a large organization, we want to make sure anything that we build is scalable and consistent,” says McMahon. “On AWS, we have standardized our approach to data using a consistent language and framework, which can be rolled out across different use cases.”

NatWest Group has adopted a number of features on Amazon SageMaker to streamline its ML workflows with the security and governance required of a major financial institution. In particular, NatWest Group adopted Amazon SageMaker Studio, a single web-based visual interface where it can perform all ML development steps. Because Amazon SageMaker Studio is simple to use and configure, new users can quickly set it up and start building ML models sooner.

To equip its data teams with the skills that they need to use these tools, NatWest Group has encouraged its employees to embark on cloud learning journeys. It has hosted over 720 AWS Training courses for its data science teams to learn new skills, such as applying best practices for DevOps and building a data lake on AWS. Additionally, several employees obtained AWS Certifications, which are industry-recognized credentials that validate technical skills and cloud expertise. By offering these opportunities, NatWest Group has equipped its data science teams to build powerful, predictive ML models on AWS at a faster pace.