Public Example

HIPAA-aware patient portal on AWS

Dacoda.ai is the chat-first AI suite for architecture-first software delivery. This example shows the type of architecture pack teams can generate from a rough set of requirements.

Blueprint Ready

HIPAA-aware patient portal on AWS

A HIPAA-aware patient portal should prioritize secure patient access, appointment and messaging workflows, auditable clinical data access, and a restrained AWS footprint that a small platform team can operate. The recommended architecture uses a Next.js frontend, API services behind an application load balancer, managed identity, encrypted relational storage, object storage for documents, event-driven notifications, centralized logging, and explicit controls for PHI handling.

Model

example:dacoda-reference-blueprint

Recommended Architecture

Use a layered AWS architecture with a public Next.js application served through CloudFront and AWS WAF, an API tier running containerized services on ECS Fargate, and a private data tier using Amazon RDS PostgreSQL for patient and workflow data. Store uploaded documents in encrypted S3 buckets with scoped presigned access. Use Cognito for patient and staff authentication, KMS for key management, CloudTrail and CloudWatch for auditability, and EventBridge plus SQS for asynchronous notification and care-team workflow events. Keep the first version focused on portal access, secure messaging, appointment visibility, document handling, and audit trails before adding advanced analytics or complex clinical integrations.

Cloud Services

AWS CloudFrontAWS WAFAmazon ECS FargateAmazon RDS PostgreSQLAmazon S3Amazon CognitoAWS KMSAmazon EventBridgeAmazon SQSAmazon CloudWatchAWS CloudTrail

Assumptions

  • The first release supports portal workflows and does not directly write to an EHR.
  • The organization will execute a BAA with AWS and configure HIPAA-eligible services accordingly.
  • Patient messaging and document access are in scope; advanced analytics are deferred.
  • The platform team can operate AWS managed services but wants to avoid Kubernetes for the MVP.

Open Questions

  • Which EHR or scheduling systems need integration after the MVP?
  • What is the approved notification policy for email and SMS content?
  • How will guardians, proxies, or family access be modeled?
  • What retention policy applies to documents and audit records?