See Demo
Customer Story

"If [Spring Health] isn't running, that 30 seconds suddenly stretches out to minutes, hours... That is somebody's lifeline being stretched out longer than it needs to be."

Jack Yuan

Associate Director of Engineering, Spring Health

How Spring Health unlocked 30–40% cloud savings with a zero-downtime AWS migration

A mental health platform serving millions migrated its entire cloud infrastructure in four months, maintained HIPAA compliance throughout, and expanded from 6 to 15 environments while keeping their DevOps team flat.

Spring Health Challenges

Spring Health’s customer base doubled year over year. Their previous infrastructure provider was built for developer productivity but imposed hard limits: container caps, single-service autoscaling, and daily backups that risked up to 24 hours of data loss. As the engineering team grew 50%, the DevOps team stayed flat.

Spring Health started on bare metal, moved to a PaaS provider that offered generous developer productivity but limited flexibility, and eventually needed more control. The PaaS approach worked well for years, but as the team needed to build more products, integrate AI/ML workloads, and operate at scale, the constraints became untenable.

The challenge intensified with the rise of AI. Spring Health’s continuous care model relies on rapid iteration: building, testing, learning, and repeating across a complex landscape of legal requirements, clinical validation, and shifting regulations. Without infrastructure that provided visibility and flexibility, the team was stuck.

DuploCloud Solutions

Since working with DuploCloud, Spring Health has transformed its entire cloud infrastructure and DevOps capability.

DuploCloud guided Spring Health through a complete migration from a legacy PaaS provider to AWS while simultaneously ensuring HIPAA compliance and gathering evidence for HITRUST recertification. The migration was not just a lift-and-shift; it involved a complete architectural modernization executed in just four months.

Spring Health transitioned from a platform with hard container limits and single-service autoscaling to a fully scalable AWS environment running Aurora RDS, OpenSearch, SageMaker, and Bedrock. Infrastructure deployment is now automated and HIPAA-compliant by default, enabling teams to self-serve without waiting on DevOps.

Instead of scaling their DevOps team linearly with engineering growth, Spring Health leveraged DuploCloud’s AI-powered agents to handle incident triage, root cause analysis, and infrastructure operations autonomously. The platform replaced manual rules engines with LLM-driven automation, keeping safety guardrails in place while offloading 80% of operational decision-making to AI.

Impact for Spring Health

All of this transformation enables Spring Health to:

  • Achieve 99.9% SLA for always-on mental health services, ensuring no patient is left waiting.
  • Reduce cloud spend by 30-40% while massively expanding AI/ML capabilities post-migration.
  • Operate at a 1:50 DevOps-to-engineer ratio, far surpassing the industry standard of 1:20-30.
  • Expand from 6 to 15 environments, with data platform and AI/ML teams self-serving infrastructure deployment.
  • Reduce maximum data loss from 24 hours to just one minute with Aurora RDS.
  • Improve overall performance by 10-20%, with up to 2x gains for specific services like OpenSearch.

More Customer Stories

Industry leaders choose to Build with Duplo

Join hundreds of enterprises managing their cloud infrastructure with Duplo. Clear your DevOps backlog and save operational costs.

Launch Sandbox