Case Study

HealthTech ISV Turns Manual Benefit-Form Mapping into an AI-Native Enrollment Engine with Claude on AWS Bedrock

At a glance

A HealthTech ISV helps hospitals connect uninsured patients to Medicaid, charity care, and 1,000+ other public assistance programs. Every benefit program uses its own form, and a single mismatched field can delay or derail a patient’s application, deepening hospital uncompensated care that industry data puts near $41.6 billion a year.

NMD built an AI-native system on Claude via AWS Bedrock that recognizes form fields across PDF and web-based applications, maps them to standardized data expressions, and auto-populates and submits completed applications, replacing manual, per-form configuration with a single reusable engine.

The ISV now has a self-extending forms infrastructure it can point at new benefit programs and hospital partners without rebuilding mapping logic each time, positioning it to scale enrollment support faster across its national footprint.

Industry

Use Case

Chatbots & Virtual Assistants, Agentic AI, Document/Form Processing

Solution implemented

The value equation

Company Snapshot

A HealthTech ISV that automates benefits eligibility screening and enrollment for hospitals and health plans nationwide, helping reduce uncompensated care and reach more eligible patients.

Location

United States

Customer Situation

This ISV exists to close a stubborn gap in American healthcare: patients who qualify for Medicaid, SNAP, charity care, or other public assistance often never receive it, not because they are ineligible, but because the enrollment process itself is broken. Hospitals absorb the cost as uncompensated care, an industry-wide problem estimated at tens of billions of dollars annually, while eligible patients delay care they cannot afford.

Every benefit program specifies its own form, its own fields, and its own submission channel. As the ISV added programs and hospital partners, its team faced growing manual work to map new forms to patient records accurately, work that directly gates how many patients can be served and how fast new programs can go live.

This is an industry-wide problem: a large share of the uninsured population is likely eligible for assistance they never access, in part because enrollment infrastructure has not kept pace with program complexity.

NMD Solution

NMD reviewed the ISV’s form-processing workflow and found the strongest leverage point was field-level intelligence: reliably recognizing and mapping form fields, rather than treating each new form as a one-off build. The solution required an AI model capable of interpreting varied, unstructured form layouts and matching them to standardized data expressions, deployed via Claude on AWS Bedrock with serverless orchestration through AWS Lambda and API Gateway, and persistent mapping and patient data storage in Amazon DynamoDB and Amazon RDS for PostgreSQL.

What We Delivered

NMD delivered a sequential set of capabilities that together form a single, continuous forms-automation engine. The system recognizes and tags fields across PDF benefit applications, maps them to the ISV’s standardized data expressions, and populates completed forms directly from patient records. That capability was then extended to dynamic web-based applications through a browser-based automation layer, so the same underlying AI logic that fills a PDF can now complete and submit an online application. The result is a reusable, AI-native pipeline the ISV can apply to new benefit programs and partners without re-engineering form logic each time.

Ready to Transform Your HealthTech Business?

See how New Math Data can transform your healthTech business with AWS-powered innovation.