Senior Software Engineer

ComPsych is modernizing its technology organization through customer-facing web and mobile experiences, a unified design system, and AI-powered product capabilities. The Senior Software Engineer will lead system design, task decomposition, AI-assisted implementation, testing, security, reliability, observability, production operations, and coaching engineers on effective AI development practices.

Responsibilities:

• Plan and design as your core contribution. You own problem framing, sequencing, and tradeoffs with Product and Design, and the system design that follows — the data model, the API contracts, the failure modes
• Direct implementation rather than writing it line by line. Decompose the work, write the specs, and let AI agents build — dropping into the code yourself when debugging or complexity demands it
• Own testing as a design decision. You decide what needs coverage and at what level, direct AI to build it, and validate that the tests prove what they claim — not just that they pass
• Build security, reliability, and observability in from the start. They're design inputs, not production afterthoughts — the systems you ship can be operated safely and understood when they fail
• Review everything with the same rigor, regardless of author, and keep production judgment human. You're accountable for what ships whether you or an agent wrote it. Deployment calls, on-call response, and what an incident teaches you stay yours
• Raise the bar for how the team works with AI, coaching engineers to specify, delegate, and validate rather than hand-code everything themselves

Requirements:

• Deep server-side experience in a modern language such as Java, Python, C#, Kotlin, or Go
• Must have 5+ years of full stack development experience
• APIs and data models other people depend on: REST contracts, versioning, OAuth2/OIDC, schema design, and SQL that holds up under real load
• Async and event-driven systems in production — Kafka or comparable, idempotency, and a clear view of what at-least-once costs you downstream
• You've built and operated production systems at real scale on AWS or Azure — containers, Kubernetes or serverless, IaC, and CI/CD pipelines you've built, operated, or materially improved. You've carried the on-call load for them, so you know what “good” looks like independent of who or what wrote the code
• A testing strategy you can defend — what belongs at unit, integration, and contract level, and how you know a release is verifiable before it ships
• Reliability and security you've been accountable for: SLOs you've been paged against, incidents you led, OWASP and threat modeling applied — and PHI/HIPAA or comparable regulated-data experience
• Agentic AI is your primary development surface, and you catch what it gets subtly wrong. Not occasional autocomplete, not a hunch — you've shipped and operated real production systems working this way, and can point to what you caught before it shipped
• LLM application engineering (RAG, retrieval quality, evals, LLM observability), or modern TypeScript frontend work against a design system