Healthcare Product Architect (AWS, SAAS)

Job Title: Lead Product Architect (Healthcare client)

Location: Remote in the USA

Contract: 12 months

Requirements

• Product architecture track record. Have led the architecture of at least one SaaS product through multiple major releases. You can talk through specific architectural decisions you made, what they cost, what they enabled, and what you would do differently now.
• Domain modelling depth. You have done serious domain modelling work, and you understand the difference between a model that reflects how the business actually operates and a model that just reflects last week's screen. You can defend a model to clinicians and operators without losing them with technical terms.
• AWS product architecture expertise. Expert-level command of AWS as a product platform across the three dimensions that matter most for Client: scalability, security under PHI and HIPAA, and performance. Scalability means you have made the production calls on Aurora sizing, Lambda concurrency, ElastiCache layering, and how the architecture absorbs load when a hundred facilities come online in the same quarter. Security means PHI and HIPAA from the ground up: VPC design, KMS and field-level encryption, IAM, Cognito, GuardDuty, CloudTrail, BAA scope, and what makes an architecture defensible in a real compliance audit rather than just BAA covered on paper. Performance means latency budgets that hold up under real clinical workflows, cold-start tuning, a cache strategy, and the discipline to actually measure rather than assume. Hands-on production experience across Lambda, API Gateway, EventBridge, SQS, ElastiCache, CloudFront, and the surrounding AWS surface area is assumed, not optional.
• TypeScript and Node.js depth. You have shipped production systems in TypeScript, and you can hold a long technical conversation about generics, inference, and type design without reaching for examples. Node.js is your daily runtime, and you understand how the event loop, async patterns, and memory profile behave under load.
• PostgreSQL at scale. You have run PostgreSQL in production, written real SQL, tuned slow queries, and managed schema evolution without breaking things. Aurora Serverless v2 specifically is a plus, but the underlying PostgreSQL skill matters more than the managed service.
• API design. You can design REST and FHIR-style APIs that hold up under real client load and integration partner scrutiny. You understand resource modelling, versioning, idempotency, authentication, and the actual mechanics of HTTP. Bonus if you have run an external API that third parties built against and survived.
• React frontend. Modern React with real production experience. You understand state management trade-offs, bundle size, and the runtime cost of the choices you make. React 19 with Vite is the current stack, and you should be comfortable with it within a week.
• Infrastructure as code. You write your own IaC. AWS CDK in TypeScript is the stack here. Terraform, Pulumi, or comparable production experience translates. ClickOps infrastructure is not part of the role.
• HIPAA literacy. You have shipped software in a HIPAA environment, and you know what that actually means day-to-day. Audit trails, BAA scope, encryption at rest and in transit, access controls, and the limits of what you can do with PHI in logs, in support, and in non-production environments.
• Architectural leadership. You have set the technical direction that other engineers have followed, written the documents that anchored long-lived decisions, and mentored engineers throughout the work. You have specific examples, and you can talk about both the wins and the misses.
• Atlassian fluency. Working knowledge of Jira and Confluence.
• AI tool fluency. Daily use of AI for code, design, and review. Anthropic Claude in particular.

Preferred

• EHR or EMR architecture experience. Prior architectural work on an EMR, EHR, or system that integrates with one. Bonus if you have shipped to post-acute, skilled nursing, or assisted living specifically.
• Regulated SaaS architecture. You have architected products through HIPAA, SOC 2, HITRUST, ONC, or comparable regulated environments. You know how compliance shapes architecture from day one rather than getting bolted on at the end.
• FHIR R4 depth. Direct production experience with FHIR R4 resources, bundles, search parameters, and the actual mechanics of building an FHIR-native system rather than an FHIR facade over a legacy schema.
• SMART on FHIR. Worked with SMART on FHIR scopes, OAuth flows, and the operational reality of third-party FHIR app launches inside an EHR.
• HL7v2 integration. Hands-on experience with Mirth, Rhapsody, or comparable interface engines. You know that real healthcare integration is messy, and you have the scars to prove it.
• ONC certification work. Time spent on an engineering team that took a product through ONC certification, or close enough to that work that you know what the lift looks like.
• Modern TypeScript stack experience. Production work with Hono, tRPC, Drizzle ORM, or comparable modern TypeScript tooling. We can ramp up on the specific tools, but prior fluency accelerates everything.
• Multi-tenant SaaS architecture. Production experience with PostgreSQL row-level security, pgPolicy, or comparable tenant isolation patterns. Bonus for healthcare multi-tenancy, where the isolation requirements are stricter than typical B2B SaaS.
• Observability and audit. OpenTelemetry, Sentry, Pino, structured logging, and distributed tracing across a serverless system. CloudWatch and CloudTrail at production depth. You have built dashboards that engineers actually use, not just check boxes.
• Testing pyramid fluency. Vitest, Playwright, fast check, Testcontainers, k6, or comparable tooling at each layer. You know how to make tests fast enough that the team actually runs them, and how to keep coverage honest rather than performative.
• Security tooling and patterns. Working knowledge of Semgrep, OWASP ZAP, CDK nag, KMS field-level encryption, and ABAC versus RBAC trade-offs. You think about security continuously, not as a checklist item at the end of a sprint.
• Real-time and event-driven systems. Production experience with WebSocket, distributed locking patterns, EventBridge, and event sourcing. The system has real-time clinical workflows, and the eventing layer matters.

Logistics

• Location. Fully remote within the United States.
• Travel. Occasional travel to facility partner sites for stakeholder meetings, integration testing, and on-site discovery. Plan on a few trips a quarter, with some coming up on short notice when a deployment needs hands in the building.
• On call. You participate in the on-call rotation. The system runs in real facilities, and engineering ownership extends to production reliability around the clock.
• Reporting structure. Reports to the CEO. Partners daily with product, engineering, and client stakeholders.