Insurance Policy Documents Email Service Modernisation with AWS Serverless Architecture
76% Run Cost Reduction and 46% Decrease in TCO
About the Customer
The customer is a British multinational insurance company that provides insurance, wealth, and retirement products to approximately 25 million customers across the United Kingdom, Ireland, and Canada. As a diversified insurer, the company offers a wide range of products for individuals and corporations, including: Life insurance and
protection plans Home, motor, travel, and health insurance Pensions, savings, and investment products.
Working with the customers’ Health Team, fourTheorem undertook a strategic initiative to modernise their REST Email Service. This is a critical service used to email policy documents to Health customers on New Business, Renewals and Endorsements. This modernisation project, executed in August 2025, aimed to reduce costs, improve scalability, resilience and enhance operational efficiency.
The Challenge
The company’s Health Team relied on traditional always-running servers (Amazon EC2 instances) for their REST Email Service, which processes an average of 800 policy documents daily for Health customers with a peak volume of 5000. The existing solution requires keeping minimum 3 (scaling up to 6) servers running 24/7, resulting in annual infrastructure costs of $1776 and a Total Cost of Ownership (TCO) of $30,346 per year.
Key challenges with this approach include:
- The architecture incurred continuous infrastructure costs regardless of usage levels. This meant that the customer was paying for compute resources even during periods of low or no usage, as a minimum of three servers were always running.
- The TCO was due to the operational overhead, including regular patching, upgrades, and management of the JBoss application server and EC2 instances.
- The application has limited scalability confined by fixed server infrastructure.
- The absence of modern infrastructure-as-code capabilities made deployment and disaster recovery procedures complex and time-consuming.
- Additionally, the shared infrastructure model increased a possible blast radius from other services that could potentially impact stability.
Current Architecture
- Java Spring application
- JBoss application server
- 3 Amazon EC2 instances (scalable to 6)
- Shared infrastructure environment
The Solution
During solution design it was recognised that the workflow is primarily asynchronous batch processing. A combination AWS Step Functions (serverless workflow service), AWS Lambda (serverless compute) and S3 (serverless object storage) is well suited to this type of problem. This insight led to the development of a fully serverless architecture that eliminated the need for always-on infrastructure while improving scalability and resilience. The architecture was transformed from an always-on legacy JBoss application on EC2 to a modern AWS Serverless architecture that can scale to zero when not in use and scale up automatically during peak processing times, paying for only the resources consumed.
The solution introduced an event driven architecture with parallel processing capabilities, replacing the traditional API interface while preserving core business logic. Infrastructure as code was used for deployment, creating comprehensive CI/CD pipelines. A complete serverless architecture was implemented using leveraging multiple AWS services:
- Amazon S3 storage for batch files
- AWS Step Functions for workflow management
- AWS Lambda for serverless computing
- Amazon SNS for error reporting
- Integration with Datadog and Splunk for observability
Results
Security Enhancements
- Implemented principle of least privilege through minimal IAM permissions
- Improved security isolation through independent deployment capabilities
Financial Impact
- Annual run cost cut by 76%
- Annual Total Cost of Ownership (TCO) cut by 46%
Operational Improvements
- Achieved “Scale to Zero” capability, eliminating idle infrastructure costs. The application can process the average daily volume that this system currently receives (800 records) within ~5 minutes with the new architecture. This means the customer will pay for only ~5 minutes of compute per day for the average volume instead of having servers on for 24 hours.
- Reduced maintenance overhead with no server patching/upgrades or application server management due to fully managed serverless services.
- Improved scalability as the system can scale up and down on demand and is not limited by previous server constraints.
- Faster deployments and enhanced disaster recovery through single-click deployment with Infrastructure as Code.
- Enhanced system resilience through native multi-AZ distribution, while independent deployment architecture minimises the impact of potential service disruptions.
About the Partner
fourTheorem is an AWS Advanced Tier Services Partner specialising in cloud-native application development and serverless architectures. With deep expertise in financial services modernisation, fourTheorem helps enterprises migrate mission-critical workloads to AWS using cutting-edge serverless technologies. The company’s senior cloud architects and serverless specialists work closely with AWS product teams to solve complex technical challenges and deliver transformational business outcomes. fourTheorem holds the AWS ECS Service Delivery specialisation, and AWS Lambda Service Delivery specialisation, demonstrating validated technical expertise and proven customer success in serverless and container-based solutions.