It is no small feat for an academic medical center’s engineering team to launch an initiative to move protected health information to the cloud. The transition is delicate given the importance of maintaining HIPAA compliance and data security. Conventional wisdom states that protected health information is more securely hosted within on-premises servers.
But Duke Crucible, a software engineering team in the Duke University School of Medicine, isn’t bound by conventional wisdom. The group is tasked with increasing Duke’s capacity for innovative health data-science development. Their mandate is to explore new technologies that can then be integrated into the greater academic health community. Committed to a cloud-first directive, the team sought out the expertise required to tackle Amazon Web Services’ complex cloud procedures. The project required not only experience with the complexities of the development but also a deep respect for the criticality of protecting health information.
Duke’s enterprise IT development team partnered with Caktus developers to build out the required technology. Together, the teams navigated and tested a variety of HIPAA-compliant AWS services to construct the overall architecture.
Through this process, the Caktus developers ensured the data was 100 percent secure. They then built an infrastructure as code (IaC) system that allows AWS cloud managed resources to be easily and rapidly deployed.To allow the data to flow seamlessly, the team developed code that integrated with the health system’s clinical networks and its continuous integration/continuous delivery pipeline. This enabled integration into both the flow of the data and the developer process using AWS CloudFormation templates and Kubernetes deployment.
The initial scope allowed for Caktus to deploy one application to the cloud as a proof-of-concept. The rapidly repeatable IaC approach then opened the door for deployment of two additional applications. The Caktus team built a nimble, scalable hosting and deployment system that will allow the health system’s developers to easily add capacity as use of their applications increases.