Challenge:
APrime worked with SAFE Health to support a fundamental expansion of their mission and scope over the past several years. Early in the COVID-19 pandemic, SAFE shifted from a smartphone app that facilitated easy, secure and verifiable STD test results to a an at-home testing app for COVID-19; more recently, SAFE built on the momentum of this transformational product to expand their vision: to reduce the cost of routine care via digital delivery of services, at scale.
APrime partnered with SAFE to design and implement the foundation of their next-generation digital health app ecosystem: a “Digital Health and Connected Diagnostics” platform that enables rapid implementation of specialized, auditable AI-enabled digital healthcare applications to facilitate remote point-of-care diagnostics.
Establishing a HIPAA-compliant point-of-care diagnostic platform posed several technical challenges, even beyond the need to achieve total compliance around privacy and security requirements. Chief among those challenges were:
- Designing the platform to be accessible to SAFE’s internal engineers as well as external partners using any programming language of choice (e.g. Go, Java, Python, and Javascript).
- Defining the protocol for how the systems communicate with each other, while building in flexibility to accommodate novel diagnostic test applications and match the fast pace of innovation in the healthcare space
- Prioritizing performance and elasticity to enable rapid deployment of testing and deployment at pandemic-level scale
To complicate matters, SAFE needed to build and deploy this new platform while actively supporting their v1 product. With existing users and customers requiring uptime and continuous release of new features, this project was a classic example of building the plane while it was already in flight.
Solution:
One of APrime’s main efforts on this project was the design and implementation of the autogeneration of language-agnostic APIs. This ensured that the SAFE engineering team, and ultimately third-party developers, could easily implement new features using the API interface without worrying about foundational design decisions.
Getting into the details:
- To ease developer adoption, we built scalable and efficient microservice SDK libraries leveraging gRPC to build language-agnostic APIs and apps in Java, Go, Python, and Javascript.
- The system leverages an event-driven microservices architecture with standardized Protobuf schema. Message validation and authorization happen under the hood, so developers only need to concern themselves with business logic.
- Microservices can interact via the NATS messaging system and also orchestrate more complex workflows with the Zeebe workflow engine from Camunda.
- Given the sensitive nature of personal health data, authentication and authorization is of utmost importance. APrime created a solution utilizing Google’s Identity service in conjunction with Authzed that is flexible, envoy-integrated, and gRPC native.
- APrime laid out foundational services within the new platform including auditing (Dataset), billing (Chargebee), and flexible configuration tracking.
- Coordinated deployment of a new class of applications on ControlPlane, a multi-cloud platform based on Kubernetes and Envoy that provides the infrastructure backbone for operating SAFE systems at scale.
Transformation:
APrime is working closely with SAFE Health’s CTO to realize his vision of this next-generation health application ecosystem, a critical component for the future trajectory of the company.
SAFE has already leveraged this new technology to successfully attract new strategic partners to its platform, and APrime’s rapid prototyping and development continue to move SAFE closer to its mission of launching a digital health and connected diagnostics platform that increases accessibility and reduces the cost of routine care at population scale.
Let Aprime help you overcome your challenges
and build your core technology
Are you ready to accelerate?