Empowering Financial Agility: AI-Driven Debt Management Modernization for the US Federal Agency

Background 

The U.S. Department of Health and Human Services (HHS) Program Support Center (PSC) provides centralized financial management services to multiple federal agencies, including the Administration for Children and Families (ACF), Centers for Disease Control and Prevention (CDC), Food and Drug Administration (FDA), and National Institutes of Health (NIH). A critical component of this support is the Debt Management System (DMS), which enables PSC to track, manage, and collect debts tied to federal grant programs and financial obligations.

Originally developed on legacy infrastructure, DMS had become inflexible, relying on paper-based workflows that introduced inefficiencies, limited visibility, and increased compliance risks. PSC partnered with ASSYST to reimagine DMS as a cloud-native, AI-enabled, secure shared services platform capable of meeting the evolving needs of multiple agencies while delivering consistent, high-quality user experiences.

Challenges 

PSC’s modernization goals were shaped by the following challenges common to legacy federal financial systems:

• Legacy Technical Debt: Outdated codebases and unsupported components hindered improvements and introduced security risks.
• Manual Processes: Labor-intensive workflows caused data errors and slowed down processing.
• Siloed Systems: Limited interoperability, restricted real-time visibility, and accurate cross-agency reporting.
• Lack of Predictive Intelligence: The system was unable to support risk scoring, delinquency tracking, or automated payment plans.
• Scalability and Mobility Gaps: The platform was unable to scale with demand or support remote and mobile users.
• Rising Cybersecurity Mandates: DMS needed to comply with stringent requirements such as FedRAMP, FISMA, and NIST.

These multifaceted issues underscore the need for a scalable, secure, and intelligent platform.

Solution

ASSYST implemented a comprehensive modernization strategy centered on agile delivery, cloud-native infrastructure, and AI-powered automation:

• Agile and Human-Centered Design: A Dedicated Scrum team collaborated with PSC stakeholders through iterative sprints. Using user personas and journey maps, ASSYST designed a responsive and accessible UI tailored to debtors, auditors, program staff, and financial portfolio managers. Features were prioritized based on value and complexity, ensuring the timely delivery of mission-critical capabilities within a 12-month timeframe.
• Cloud and Infrastructure Modernization: The legacy DMS was migrated to a secure AWS environment. Dedicated cloud environments were provisioned for development, staging, and validation, with the application deployed using Angular for the frontend and Node.js for the backend. Built on a microservices architecture and containerized using Docker, the solution leveraged custom Kubernetes clusters on Amazon EC2 instances, utilizing kOps for orchestration. This enabled rapid scalability and modular enhancements. RESTful APIs facilitated integration with payment processors and mission-critical systems. A fail-safe, automated cutover plan ensured zero downtime during the transition.
• DevSecOps and CI/CD: A comprehensive DevSecOps pipeline was implemented using Jenkins, AWS CodeCommit, and SonarQube, integrating static code analysis, security scans (OWASP), and compliance checks into each sprint cycle. The CI/CD process automates build, test, and deployment across Kubernetes clusters running on EC2 instances in the DEV, STAGING, and VALIDATION environments, all managed via kOps. Docker images were built and stored in Amazon ECR, then deployed as pods within the Kubernetes clusters. Access control was using AWS IAM, with secure VPC peering connecting all isolated environments. Secrets and credentials were managed through AWS Secrets Manager, while centralized monitoring and alerting ensured operational visibility and system health across all stages of the deployment pipeline.
• AI and Intelligent Automation: AI/ML models were integrated to make predictions and extract text from debt documents, and the Automation of the new debt creation process was achieved through an integrated, trained language model. Automated services handled repetitive tasks such as debt record imports and workflow initiation, reducing manual errors and increasing processing speed.
• Advanced Analytics: Developed interactive, high-performance reports in AG-Grid and MySQL, designed to handle large datasets efficiently, and offered a wide range of features to visualize and manage tabular data that were accessible via role-based dashboards.
   

Results 

The modernized DMS platform has delivered a measurable impact:

• Scalable Shared Services: Supports onboarding of new agencies and increasing data volumes without degradation.
• Real-Time Financial Visibility: Live dashboards allow agencies to optimize strategies based on current conditions.
• User-Centric Access: Mobile-ready self-service portals streamline workflows and improve stakeholder satisfaction.
• Strengthened Security and Compliance: Automated controls and continuous monitoring ensure adherence to federal mandates.
• Resource Optimization: Automation of routine tasks reduces operational overhead and boosts productivity.
• Enhanced Interoperability: Secure, role-based multi-agency access fosters collaboration while maintaining data separation.

Sustaining Growth and Innovation

The platform modernization delivered by ASSYST provides PSC with the agility to evolve with future demands. With cloud-native flexibility, AI-enabled automation, and continuous delivery pipelines, PSC is equipped to onboard new debt programs, meet changing regulatory mandates, and deliver greater transparency in debt management.

ASSYST’s solution has empowered PSC with the tools, architecture, and operational model to sustain and scale the DMS platform as a strategic shared service, enabling data-driven decision-making and long-term mission success.