AWS Transformation Improves Health of Regional Blues Plan

Modernizing the data platform

Already using AWS in other parts of the business, the insurer knew that it wanted to move to an AWS-based data lake with the latest technologies. Specifically, it possessed a plethora of structured claims data housed in a relational structure that hindered its ability to use unstructured data from a variety of sources, including the member portal and member encounters. To accomplish the organization’s goal to use the unstructured data coming into the environment, absorbing, and correlating it to find areas to improve patient satisfaction and support its value-based care programs, the two teams began work building the new data lake environment.

Nucleus accelerator

The Nucleus Intelligent Enterprise Platform uses NTT DATA and third-party IP to accelerate deployments; for this project, the team relied on Nucleus’s data and intelligence fabric to help accelerate the data lake build. Nucleus combined with NTT DATA consultant expertise in AWS technologies meant the team was able to get the data lake framework established within four months.

Nucleus accelerators included data ingestion pipelines that added further value to the AWS data lake. Data ingestion pipelines were created using Apache Spark; two ingestion modalities were created, using Scala and Python with AWS Simple Storage Service (Amazon S3) buckets as the destination for the data. The solution defined and implemented data lake zones to which data could automatically be directed.

Zones included a raw or landing area layer for data persistence and retention of raw data assets; a cleansed or curated layer, where transformed data is available in consumable data sets; and a production or secure layer where business logic and row level security are added. From these layers, data ends up in a presentation layer where line-of-business customers can gain access.

For data processing, the solution relies on Amazon EMR, a big data platform that using Apache Spark can process the insurer’s large amounts of data. With Amazon EMR, the insurer can run its workloads on Amazon EC2, or Amazon Elastic Kubernetes Service (EKS) clusters, which it plans to expand to as it matures its data lake program.

The implementation is deployed using AWS CloudFormation templates to help ensure configuration consistency and includes integrations with GitLab and GitHub.

Security and compliance built-in

Amazon EMR not only automates tasks like provisioning capacity and tuning clusters, but it is also inherently HIPAA compliant. Beyond EMR’s HIPAA compliance, the team built-in security by installing certificates to ensure communications had correctly defined security protocols, including data at rest and data in motion. This included data within Amazon Virtual Private Cloud (Amazon VPCs) where security protocols were proactively built in. NTT DATA consultants encrypted protected health information using Kerberos-based authentication.

Additional security measures included Amazon S3 encryption, Atlas authentication, and more. In addition, NTT DATA consultants established monitoring with AWS CloudTrail and provided the client with the ability to detect, respond, investigate and remediate incidents using logs from Amazon S3 and Amazon VPC. Amazon CloudWatch provides the team with visibility into system performance and health.

With a solid data lake platform in place, the NTT DATA team handed the reins to the client for data population. As the customer migrates data to its new AWS data lake, NTT DATA experts remain involved, ensuring the client team has the knowledge needed to effectively migrate data to the new system and maximize the new system capabilities.

Benefits

As a data-centric organization, this health plan now has a platform at its fingertips to begin driving data insights to further its goals of providing more personalized care and fostering collaboration among providers. By bringing together healthcare claims data from multiple processing systems, and unstructured sources like patient touchpoints, this regional health plan now has the power and flexibility to build personalized experiences that will boost loyalty and entice new members as well.

The modernized data lake grows the company’s efficiency over 33% from its expensive, on-premises legacy platform. By separating the data and compute layers and moving the firm’s data to Amazon S3, the insurer was able to drastically decrease its storage costs. Moreover, by replacing several expensive tools with open-source solutions, the company was able to save on execution costs as well.

Now the health plan’s internal business customers have latency-free access to critical data sets that will allow them to bring new data-proven ideas to market faster.

Next steps

NTT DATA continues to work with the client, ensuring its platform, related applications, tools and development mature as desired. Planned steps in the process include expanding past EC2 to use managed Kubernetes containers and EKS as well expanding the use of Lambda functions. Last, the insurer plans to adopt AI and ML, creating advanced data models for even deeper patient insights.