Driving Efficiencies in Senior Market Insurance with AI & Data

Wellcove (previously CHCS Services) has administered senior insurance programs for over 25 years. Based in Pensacola, Florida, the company manages long-term care, Medicare Supplement, life, disability, and supplemental health programs. It serves carrier partners across the United States.

01 The Challenge

A growing book of business that called for a single view of operations

Health insurance fraud costs the U.S. an estimated $36.3 billion per year. Medicare alone loses roughly $60 billion annually to fraud, waste, and abuse. For administrators like Wellcove, the ability to catch suspicious patterns early can save carriers millions. Doing that well requires clean, centralized data.

Wellcove’s data lived in separate on-premises systems. Each department, claims, call center, finance, worked from its own data sources. The setup had served the company for years. As the book of business grew, leadership saw an opportunity to bring everything together. A centralized view would strengthen SLA reporting and speed up contract negotiations. Executives would get real-time visibility into revenue and inventory trends.

The team also saw a clear use case for AI. Fraud detection across the full claims dataset would be possible once the data sat in one place. The existing on-prem setup was not designed for that kind of workload. Wellcove wanted a modern data foundation on AWS. AI fraud detection would be the first application on top of it.

Wellcove connected with Provectus during an AWS Gain Insights session. After an initial review, Provectus scoped the engagement in two parts. Build the data platform, then deploy the fraud model.

02 The Approach

Map every data source, then design the platform around reporting and compliance needs

Provectus started with discovery. The team investigated Wellcove’s data sources: SQL engine versions, access requirements, SFTP needs, data availability, and overall volume. They documented reporting requirements for every dashboard. Custom calculations, chart types, data freshness SLAs, and source table schemas.

Privacy, security, and compliance got their own assessment. Insurance data carries strict handling requirements. The architecture needed role-based access controls and audit trails from day one.

From discovery, Provectus defined the data warehouse design. Which Wellcove clients to serve first. Which data sources to connect. Which reports and dashboards to prioritize. The fraud detection model would sit on top of the same data layer.

03 The Build

Cloud data lakehouse, automated pipelines, Power BI dashboards, and AI fraud detection

The data platform was built as a cloud data lakehouse on AWS. Automated pipelines ingest and process data from Wellcove’s source systems into a centralized data lake. A data catalog makes datasets discoverable. Quality controls and consistency checks run on every dataset before it reaches analysts or dashboards.

Role-based access controls manage security and governance across the platform — required for compliance-heavy insurance data.

On top of the data layer, Provectus deployed Power BI dashboards covering:

• SLA performance across clients and contracts
• Revenue and inventory trends by segment
• Claims department operations and metrics
• Call center performance and customer service KPIs
• Finance department reporting

The final deliverable was the AI fraud detection model. It runs against the centralized dataset. It identifies patterns across claims that manual review would miss or catch too late.

04 The Results

Centralized analytics and AI fraud detection, live in six months

Provectus delivered the full platform in six months. Data lakehouse, automated ingestion, dashboards, and fraud detection. Wellcove executives now see SLA performance, inventory levels, and revenue trends across the business in one view.

6 Mos.

From siloed on-prem data to production analytics and AI fraud detection

All departments now share a single data source. Decisions that used to start with manual data pulls now start with a dashboard. The AI fraud detection model flags suspicious claims patterns across the full dataset automatically.

For Wellcove’s carrier partners, the impact is direct. Faster fraud detection reduces claims leakage. Better SLA visibility strengthens contract negotiations. Centralized reporting lets Wellcove demonstrate operational performance with data.

05 What’s Next

A data foundation built for AI, with fraud detection as the first use case

The fraud detection model was Wellcove’s first production AI deployment. The data platform was designed to support more. Provectus works with Wellcove on additional AI use cases across claims processing, customer service, and care management.