Healthcare System Migration & Performance Optimization

HOME – Case Studies – Healthcare System Migration & Performance Optimization

Healthcare System Migration & Performance Optimization

Migrated a hybrid serverless/monolithic architecture to a scalable microservice design, moving from DynamoDB to RDS and reducing AWS costs by 35% while improving API performance 3×.

Case Study

Modernizing a Hybrid Serverless Healthcare Platform

A U.S based healthcare provider managing AED (Automated External Defibrillator) devices and First-Aid stations was operating a hybrid backend — combining serverless AWS Lambda functions (for the mobile app) with a Spring MVC monolithic backend (for the web portal and cache worker).
While the system worked functionally, this combination introduced significant architectural inefficiencies, inconsistent behavior between mobile and web, and high AWS costs.
Vivid Soft led a comprehensive modernization project that migrated the data layer from DynamoDB (NoSQL) to Amazon RDS (MySQL), consolidated key backend functions into a Spring Boot microservice, and designed a roadmap to unify the web and mobile systems under a single API layer.

The Challenge

The legacy architecture had evolved organically, mixing paradigms that clashed as scale increased.

Key Pain Points

Hybrid Complexity: Mobile APIs served by Lambda + API Gateway, while web and cache operations ran on a Spring MVC monolith. Both contained duplicated business logic.
Data Model Misfit: Highly relational data (devices, inspections, supplies) was forced into DynamoDB, requiring complex joins simulated in code.
Cache Overhead: A cache worker executed cron jobs every 15 minutes to fetch, aggregate, and store data into ElastiCache, producing stale data and heavy resource consumption.
High AWS Bills: DynamoDB reads/writes, Lambda invocations, and cache storage collectively drove infrastructure costs.
Inconsistent Behavior: Feature parity between mobile and web was difficult to maintain due to code duplication.

Our Approach

Our Solution

Vivid Soft implemented a multi-phase modernization plan that balanced execution and strategic redesign.

Database Migration

Designed a normalized MySQL schema in Amazon RDS tailored to the application’s relational nature.
Utilized Flyway for schema versioning and controlled releases.
Developed temporary data migration Lambdas to extract and transform data table-by-table from DynamoDB to RDS.
Achieved zero downtime migration, ensuring live system availability throughout.

Backend Consolidation

Replaced several Lambda-based endpoints with a consolidated Spring Boot microservice, deployed on AWS Elastic Beanstalk.
Leveraged Spring Data JPA (Hibernate) for efficient ORM-based data access, removing the need for pre-aggregated cache layers.
Simplified caching strategy by retaining only targeted caches where latency benefits justified them.

Frontend & API Unification

Infrastructure & Automation

Outcomes

Results

Metric

Before

After

API Response Time

~1.8 s

~600 ms

AWS Monthly Spend

100% baseline

–35%

Cache Refresh Cycle

15 min scheduled cron jobs

On-demand queries

System Availability

Inconsistent

99.9% uptime

Development Overhead

Redundant logic (web + mobile)

Single microservice codebase

Business Impact

AWS costs reduced by consolidating DynamoDB, ElastiCache, and Lambda workloads.
Data retrieval times improved by nearly 3× with RDS and optimized queries.
Simplified development lifecycle with unified APIs and automated deployments.
Established a scalable, maintainable foundation for future feature growth.

Summary & Key Insights

Client Details

Industry: Healthcare & Emergency Equipment Management
Region: United States
Engagement Model: Contractor

Tech Stack

Languages & Frameworks: Java, Spring Boot, Spring MVC (legacy), Spring Data JPA
Databases: DynamoDB (legacy), Amazon RDS (MySQL), Redis (selective caching)
Cloud & DevOps: AWS Elastic Beanstalk, Jenkins CI/CD, CloudWatch, API Gateway (legacy), Flyway
Security: JWT, IAM Policies, HTTPS
Architecture: Hybrid (Serverless + Monolith) ➔ Unified Microservice

Key Takeaways

Migrated from DynamoDB to RDS (MySQL) to align data model with business domain.
Consolidated redundant Lambda functions into a centralized Spring Boot microservice.
Reduced AWS costs by 35% through infrastructure optimization and simplified caching.
Proposed a unified API and frontend modernization strategy for long-term maintainability.
Established continuous deployment pipelines and real-time monitoring for operational reliability.

Key Initiative

AWS Cost Reduction

Stacked bar chart showing AWS daily cost by service component before and after migration. Major reductions in DynamoDB, Lambda, and ElastiCache.