Driving Efficiencies in Senior Market Insurance with AI & Data
Wellcove enables data-driven decision-making, optimizes operational costs, and implements ML fraud detection on its new Data Platform
Wellcove (previously “CHCS Services”) is a premier administrator of senior market and eldercare programs, dedicated to providing outstanding outsourcing solutions and personalized services to their clients. Their comprehensive administrative support is complemented by various customized care management solutions that help to control medical claims costs and enhance customer satisfaction. Wellcove understands the importance of providing exceptional care to seniors and their families, and their team is committed to delivering compassionate, top-quality service every step of the way. With Wellcove, you can trust that your loved ones are in good hands.
Wellcove wanted to start leveraging business intelligence and analytics tools on top of a modern data platform. By gaining a better understanding of key metrics such as service level agreements (SLAs), revenue, and inventory trends across different segments, Wellcove hoped to make effective data-driven decisions for negotiating contracts and optimizing staffing resources. Wellcove also wanted to provide valuable insights to their customers, enhancing the customer service experience. Achieving those goals would require a comprehensive analytics strategy and an advanced data infrastructure.
Provectus approached the project through a series of iterations, covering digitalization of the data analytics platform, minimization of manual processes in data pipelines, and introduction of new tools for effective BI & Analytics. The work involved platform discovery, solution implementation, and solution integration phases: using AWS services for CDC and ETL; setting up the infrastructure and pipelines for data quality, governance, and observability with Open Data Discovery (ODD) platform and AWS services; deployment of the data platform with all required BI & Analytics dashboards.
In six months, Provectus built a robust, scalable, and secure data platform for BI & Analytics, designed to deliver valuable insights, enhance customer experience, and position Wellcove as a market leader. The platform’s new dashboards enabled Wellcove executives to look into various SLAs, inventory and revenue reports, and access various applications. They provided Wellcove with greater insight into operations of their claims, call center, and finance departments, helping to analyze various metrics and KPIs. The first AI use case, ML fraud detection solution, was successfully implemented.
A cutting-edge Data Platform delivered in six months
A robust Data and ML infrastructure for AI adoption
An ML fraud detection solution implemented
Improved visibility into disjointed data pipelines
Data Decentralization and Limited Access to Analytics as Factors Contributing to Increased Operational Costs
Wellcove (CHCS Services) is a renowned provider of senior market insurance solutions, including administrative services, supplemental healthcare claim support, and personalized care management. With a strong focus on the senior market, Wellcove excels in risk evaluation, seamless premium management, and customer experience optimization. Wellcove is committed to delivering value for carrier partners and policyholders alike. They leverage cutting-edge technologies to ensure excellent service across various markets.
As an established business with heavily structured operations, Wellcove was looking to streamline and optimize their processes and reduce operational costs. Data — and, most importantly, business-critical insights to be extracted from it — was perceived as a pivotal catalyst for this transformation.
Getting the transformation off the ground was challenging, however, due to the company’s pervasive use of compliance-heavy insurance data, dispersed across disjointed, siloed data pipelines. With an outdated on-premises infrastructure, significant improvements were necessary to develop and implement a modern platform for handling dispersed data, often in real time. Incorporating advanced capabilities for analytics and business intelligence was critical for Wellcove teams to be able to tap into the data, to drive actual business value.
An updated data platform could also lay the groundwork for AI adoption and ML development — a long-term strategic goal of Wellcove leaders.
As their pioneering effort, Wellcove’s leadership wanted the ability to access and use insights from data. The initial goal was to improve visibility into Service Level Agreements (SLAs), and revenue and inventory trends. Understanding these patterns across various segments would facilitate data-driven decisions, encompassing contract negotiations and efficient resource distribution. Wellcove also sought to gain valuable insights about their customers, to enhance the customer service experience.
The need for a thorough overhaul of Wellcove’s infrastructure was evident.
Provectus, an AWS Premier Consulting Partner, came into the picture during Wellcove’s Gain Insights session with AWS. After an initial review, it was resolved that Provectus would assist Wellcove to enhance and operationalize their data platform solution by implementing dashboards for reporting and analytics that would be connected to a data lake on AWS.
Adding Discoverability and Observability to Wellcove’s Data Solution, and Cloud Data Lake Development
During this project, Provectus’ primary technology focus was to design, build, and implement a customized data lakehouse infrastructure. Key tasks included the development of a data lake and data warehouse; implementing deployment processes for data ingestion, lake formation, and representation. Furthermore, data pipelines were to be established, enabling the creation and utilization of various dashboards and reports. The implementation of rigorous production monitoring and maintenance was also part of the infrastructure enhancement and optimization effort. As a result, Wellcove would receive a fully functional data lakehouse infrastructure, embodying high-level functionality as outlined in the project discovery session, seamlessly deployed in Wellcove’s environment.
When delving into the specifics of the Wellcove project from a technological perspective, Provectus took into account the following:
- To reduce data latency and ensure transaction integrity during ingestion, the existing Change Data Capture (CDC) process should be scaled horizontally. Continuous data replication should also be maintained. Using AWS Database Migration Service (DMS) and Amazon S3 is recommended.
- To reduce ingestion friction for new data and to enable it to integrate seamlessly into a high-performance data warehouse, advanced data lake and data lakehouse design patterns should be applied. Using Amazon S3, AWS Glue, and Amazon Redshift is recommended for data storage.
- For moving data from source systems to a data warehouse, either ETL or ELT can be used. In Wellcove’s architecture, it is recommended to conduct basic data transformations prior to loading the data into Amazon Redshift with AWS Glue.
- Implementing Open Data Discovery (ODD) platform and data quality for data management is required for tracking data lineage and quality, to ensure data trust. Data quality monitoring should be employed to facilitate identification and resolution of data trust issues such as outdated calculations, incomplete data, job failures, and missed SLAs.
- To streamline and facilitate querying of data from a unified SQL interface, Amazon Redshift Spectrum should be used. Amazon QuickSight can be used for self-service BI Reporting, to democratize data access. AWS Lake Formation should be employed as a secure data lake, with its fine-grained access control for users, AWS services, and third-party applications.
- In Wellcove’s architecture: If ETL is built using AWS Glue, the AWS Glue workflow should be used for orchestration; if multiple AWS services are used, either Amazon Step Functions or Amazon Managed Apache Airflow are options. These services can orchestrate workflow and enable Wellcove’s engineers to look into individual workflow steps, to identify issues.
Considering these factors, the proposed solution architecture is as follows:
Provectus began by thoroughly investigating the available data sources. This included examining the versions and types of SQL Engines in use, understanding the access requirements, and identifying any Secure File Transfer Protocol (SFTP) access needs. The team also reviewed the data availability across multiple environments and assessed the overall volume of data.
Provectus investigated the reporting business requirements for every report and dashboard. This involved documenting custom calculations required for each report, identifying the types of charts used for data presentation, and outlining the data friction requirements and Service Level Agreements (SLAs). The team defined the schemas of the source data and tables used, and documented the data flow.
Provectus also evaluated the project’s privacy, security, and compliance requirements. The team examined the implications of multi-tenant vs. contractual segmentation, but purely for discovery purposes.
Upon completing the investigation and discovery phases, Provectus used the insights gained to define the MVP — a data warehouse infrastructure. This involved deciding on which Wellcove clients to serve, what additional data sources to use, and the types of reports and dashboards to create.
Implementation of the data platform solution for Business Intelligence & Analytics involved the following:
- AWS DMS was used to implement CDC pipelines for the source data
AWS DMS, AWS Glue, and AWS Lambda were used to load data into the cloud data lake
- AWS Glue Crawlers was set up to initialize the data catalog
- The ODD Platform was deployed and configured to enhance data discovery and observability
- Data Quality pipelines were implemented to ensure the accuracy and validity of data flows and transformations, preventing errors and ensuring that datasets met expectations. To validate the integrity of data, rigorous data consistency testing was conducted
- For effective data transformation, ELT data pipelines were set up. To further ensure data security and access control, AWS IAM’s roles and permissions were used
Upon successfully integrating the solution, Provectus prioritized deliverables for the selected use case and MVP — an AI/ML-powered fraud detection solution. The team deployed the fraud detection solution to the Wellcove environment.
In the final stage, Provectus set up Power BI dashboards for the data platform. This provided Wellcove with a user-friendly interface to interact with their data, to derive valuable insights about their operations, processes and customers.
Provectus Solution as a Foundation for AI Adoption, ML Development, and Advanced BI & Analytics
In six months, Provectus delivered a robust, scalable, and secure data platform for BI & Analytics, designed to offer valuable insights, enhance the customer experience, and position Wellcove as a market leader. This advanced platform significantly improved visibility into previously disjointed data pipelines, enabling Wellcove executives to monitor various SLAs, inventory, and revenue reports, and enhancing application usage through intuitive, comprehensive dashboards.
The dashboards also provided insights into the operations of the claims, call center, and finance departments. They facilitated the analysis of various metrics and KPIs, leading to more informed decision-making and efficient operations.
The solution, built using AWS services following AWS best practices, proved to be a turning point in handling Wellcove’s compliance-heavy insurance data. It enabled Wellcove to reduce operational margins while introducing AI/ML-powered fraud detection. This AI use case demonstrates the potential for AI adoption in Wellcove operations.
The MVP solution delivered by Provectus served as a proof of concept for Wellcove, providing a data warehouse infrastructure for a limited number of clients. Provectus ingested and consolidated data from separate sources and created dashboards in Power BI for SLAs and various other metrics. This gave Wellcove more visibility into the quantity of SLAs from different sources and laid the groundwork to continue business transformation.
The work carried out by Provectus has enabled Wellcove to add value to their suite of products and services through comprehensive reporting and data analytics. It also laid a foundation for further advancements in the use of data and AI/ML solutions in Wellcove operations.
Looking to explore the solution?