Adapting Not Only to Survive But to Thrive

It’s no secret the insurance industry has been slower to take advantage of cloud technology and automation strategies
than many other sectors. But that doesn’t mean customers seek to interact with an insurance company differently than
other retailers and service providers. Just as individuals are increasingly using mobile phones to conduct banking
transactions, they are also turning to digital channels to address everyday insurance needs quickly.

Legacy insurers realize the need to transform their approach to digitalization to relieve internal development
constraints that limit a developer’s ability to build, test, and deploy new products rapidly and provide customers the
insurance experience they expect. Stelligent helped one such insurance company achieve faster feedback loops for
developers by creating an automated Continuous Integration and Delivery (CI/CD) pipeline running on AWS.

Using Manual Development Processes at the Expense of Rapid Innovation

This particular insurance company provides private mortgage insurance to U.S.-based mortgage lenders; it protects
lenders from default-related losses and enables homebuyers to purchase homes more quickly with down payments less than
20 percent. Before engaging with Stelligent, the company began heavily using AWS to support their suite of mortgage
processing applications. However, the company’s build and deployment processes included many manual steps, leading to
inconsistent deployment and slow feedback loops on system changes for developers.

Lacking a consistent and automated build and deployment process led to slow release times and left a lot of the
potential benefits of running on AWS unrealized. Stelligent began working side-by-side with the company’s development
teams to help the company speed up its feedback loops for developers and ensure consistency across deployments through
automation.

Turning to Stelligent to Build an Automated CI/CD Pipeline on AWS

Stelligent, a Premier AWS Consulting Partner and AWS DevOps and Financial Services Competency Partner, began by
evaluating the steps of the build and development process the company already had automated, and manual steps in place.
To develop the CI/CD pipeline on AWS, Stelligent took advantage of many AWS services, including:

• AWS CloudFormation to script and provision AWS resources and
  node configuration in a consistent and repeatable fashion
  

• 
  Amazon Elastic Compute Cloud (Amazon EC2) for
  on-demand compute power and scalability
  
• 
  Amazon Simple Storage Service (Amazon S3) for
  large-scale data storing capabilities
  
• 
  AWS Identity and Access Management (IAM) and AWS Security Token Service
  (STS) for secure controls over user access to AWS services and resources
  
• 
  Amazon DynamoDB for access to a fast, flexible,
  and fully managed NoSQL database
  
• 
  Amazon Simple Workflow (SWF) to provision workflows and tasks
  in an automated and repeatable manner
  

The fully automated solution uses many open source tools including Checkstyle, Chef, Cobertura, AWS Software Development
Kits (SDKs), Jenkins, Jenkins Job DSL, JUnit, Maven, Ruby, and Serverspec.

Every step of the build and deployment process, including the provisioning of AWS infrastructure, node configuration,
and application code deployment, are committed as code to a version-control repository. Every change to the system can
be automatically built, analyzed, and tested.

The fully automated pipeline first provides fast feedback by applying static analysis tooling and automated tests to
each change. The testing can measure the quality of the change and whether it will cause a regression in a matter of
minutes. Second, the pipeline provides realistic feedback per commit. It builds a production-like environment from code
in a repeatable, reproducible way and affords developers the opportunity to run automated tests against it in a
relatively short amount of time.

Exceeding Customer Expectations by Moving Quickly

Taking advantage of a fully automated pipeline helps the company’s developers gain the confidence they need to focus on
adding and releasing new and improved software features to market rather than spending months pushing out one software
release. This, in turn, enables the company to drive additional business value to its customers and respond rapidly to
customer requests. Additionally, the company has realized significant cost savings through its newfound ability to
experiment, test, and push code to production in a consistent and automated fashion. Instead of each launch becoming a
substantial time investment, the deployment of software happens with the confident click of a button.