Healthcare analytics is an emerging area of data science that aims to make sense of the enormous volume of data, often unstructured and analog, generated in hospitals and clinics every day. The 2020 pandemic, however, highlighted many of the obstacles faced when sharing health data across, and data siloes within, organizations.
Circle of Life offers health institutions cloud-based analytics tools to facilitate data-driven decision-making. Its main product, ZEVAC, accesses more than 7 million patient records each day to analyze how medication—primarily, antibiotics—are used. ZEVAC is a software as a service (SaaS) currently used to process data from multiple hospitals across India.
We’ve experienced greater processing power and faster computing on AWS.”
Lead DevOps, Circle of Life
Improving Uptime with Kubernetes on AWS
ZEVAC is a containerized application that runs in Kubernetes clusters in the cloud. In 2020, the company’s cloud provider experienced several bouts of downtime that interrupted customers’ ability to interact with ZEVAC—a round-the-clock, high-availability system. That same year, Circle of Life decided to migrate the SaaS to Amazon Web Services (AWS) to improve uptime.
Circle of Life worked with AWS Partner PC Solutions to migrate ZEVAC and other peripheral applications to AWS. In the year since migration, the company and its customers have experienced zero downtime with the ZEVAC platform, with 99.999 percent availability.
Supporting Kubernetes Workloads with 200 Windows Virtual Machines
Before the migration, Circle of Life’s engineers had to manually monitor and check whether its container orchestration tool was updated when Kubernetes configurations changed. It now uses Amazon Elastic Kubernetes Service (Amazon EKS) for container orchestration and Amazon Relational Database Service (Amazon RDS) to manage PostgreSQL and MySQL databases. With Amazon EKS, the company benefits from automatic updates and version control.
ZEVAC deploys in Docker containers running on Amazon Elastic Compute Cloud (Amazon EC2) instances for Microsoft Windows Server. Currently, Circle of Life runs several hundred Windows virtual machines to support its Amazon EKS nodes. PC Solutions right-sized instances for optimal compute versus cost, and app-integrated Amazon CloudWatch stack for monitoring. When traffic exceeds the 60 percent threshold, autoscaling provisions additional resources, which ensures ZEVAC remains highly available and durable regardless of data processing volumes.
Gaining Intuitive Dashboards and 25% Faster Deployment
Since migrating to AWS, Circle of Life has received positive feedback from its external and internal customers on improved application performance. While PC Solutions was initially managing its AWS environment, Circle of Life’s IT team has since taken over and finds the AWS console simple to work with.
Pundarikaksha Mishra, lead DevOps at Circle of Life Healthcare, says, “AWS dashboards are intuitive, which allows smooth performance of any task.” He continues, “The team at PC Solutions helped with the transition to AWS, which was extremely valuable as our team was new to the platform. The support we’ve received directly from AWS has also been amazing. Within minutes of raising a query, we get a response.”
Speed has likewise improved on AWS, as deploying new instances is faster. In Circle of Life’s previous cloud environment, it took at least an hour to deploy a new instance, whereas on AWS engineers can deploy new Amazon EKS nodes in 40 minutes.
Automating CI/CD Pipeline Reduces Costs by 15%
In addition to the Kubernetes migration, PC Solutions worked with Circle of Life to automate the company’s continuous integration/continuous development (CI/CD) pipeline. Circle of Life is now using AWS CodePipeline as a fully managed continuous delivery service and Jenkins as an open-source automation server. By integrating native AWS and open-source tools, Circle of Life has reduced its costs by 15 percent.
Dhananjay Yogi, head of cloud services at PC Solutions, explains, “We successfully integrated Jenkins with Amazon Resource Names, which automatically spins up Amazon EC2 instances on-demand to run Amazon EKS clusters. All updates and patches are performed automatically without downtime or manual effort, so performance has improved while lowering cost.”
Supporting Prescription Decisions with Artificial Intelligence
Circle of Life’s roadmap for ZEVAC includes enhancing artificial intelligence to help guide physicians’ decisions when prescribing medication. The company continues to consult with PC Solutions and AWS to support evolving and potential use cases in the cloud. Mishra says, “We’re now thinking of ways to intelligently recommend the course of antibiotics for each patient based on empirical data and the patient’s profile.”