overview
Healthcare & Life Science research is being transformed thanks to recent developments in AI technologies and cloud computing. This particular case study looks at a platform native to the cloud and its enhancement of data processing. The AI enabled hybrid cloud scaling and inferencing in the solution provided gave real-time insights, which streamlined processes and increased compliance, which ensured outcomes were reached and collaboration was enhanced further.
Scope
- Cloud-Native Edge Platform System Modernization: Engineering design and delivery services to support cloud-native edge AI workloads for medical modalities.
- On-Demand Hybrid Cloud Scaling: Co-processing DICOM medical modality images in the cloud to support third-party AI processing using cloud resources.
Solution
Designed an Edge Platform AI Processing Platform with Kubernetes-as-a-Service for workload scheduling and mission-critical resiliency.
Simplified cost and resource management through an integrated model for unmanaged IaaS resources.
Established observability across workloads with deployment automation pipelines and GitOps delivery.
Provided an API-First Hybrid Cloud Management Platform for workload migration between AWS and on-prem Kubernetes.
Approach
Hybrid Edge Provisioning Platform:
- Enabled CI/CD integration into Terraform for IaaS provisioning and Kubernetes fleet management, supporting Multi-Access Edge Computing (MEC).
- Delivered real-time AI inferencing with sub-4-second response times to meet FDA 92-second SLA requirements.
- Centralized deployment of compute, workloads, and storage using a unified interface.
Shift Left Security Model
- Automated risk scoring for workloads via a DevSecOps pipeline with self-service risk approvals.
- Enforced policy-as-code using Open Policy Agent for resource tagging during deployment.
Benefits
- Improved patient outcomes with rapid remote diagnostics and system recovery.
- Generated clinical insights with advanced AI algorithms on multisource edge data.
- Enabled third-party application delivery with self-service capabilities.
- Supported 100+ services through a catalog-as-a-service model.
- Reduced field service needs with embedded remote diagnostics.
- Enhanced observability for failure analysis and actionable data points.
- Streamlined hybrid platform scaling across metro-cloud and public cloud resources.