Staff Software Engineer, AI-Native Systems

Wheel

4d•$185,725 - $264,500•Remote

About The Position

Wheel builds the infrastructure for virtual care, focusing on AI-native systems that replace manual processes with intelligent, agent-powered solutions. They are innovating in a regulated healthcare environment to create the next architecture of virtual care.

Requirements

8+ years building and operating production software, with meaningful full-stack depth across a TypeScript/Node.js backend and at least one other language (Python strongly preferred).
A track record of technical leadership as an individual contributor: owning a domain, leading multi-engineer efforts to completion, and driving decisions across team boundaries without positional authority.
Hands-on experience designing and deploying agentic systems in production — retrieval, orchestration, tool/function calling, and evaluation — with a clear-eyed view of where LLMs and agents work and where they don't.
Demonstrated ability to take a loosely defined problem and drive it to a shipped, measured, agent-powered workflow.
A working practice of using AI development tools as a force multiplier, with judgment about when to trust, verify, or override them.
Strong cloud-native engineering fundamentals; comfort with CI/CD, observability, and running what you build.
Fluency with relational data and SQL.
Clear written and verbal communication, including the ability to write a design doc that changes minds.
Comfort with ambiguity and a bias toward shipping measurable results.

Nice To Haves

Experience with agent frameworks and multi-agent architectures at production scale.
Model evaluation and guardrail infrastructure — measuring output quality, catching regressions, keeping agents inside safe bounds.
Experience building platform capabilities consumed by other engineering teams.
Background in workflow automation, forecasting-driven products, or supply-demand matching.
Prior work in a regulated environment (healthcare/HIPAA, fintech, etc.) and an instinct for the constraints that come with using AI on sensitive data.
Experience mentoring engineers or acting as a formal tech lead.

Responsibilities

Own the technical direction for a significant AI-native domain: agent architecture, platform abstractions, or evaluation and guardrail infrastructure.
Act as tech lead for a squad or a cross-team initiative — decomposing ambiguous problems, sequencing delivery, identifying and clearing blockers, and keeping the team pointed at the outcome rather than the ticket.
Write and review design docs; make and document the load-bearing architectural calls, including the ones where the answer is "not yet" or "buy, don't build."
Establish clear technical ownership where it's currently diffuse, so decisions have a named owner and reviews don't stall.
Design and build production AI agents incorporating retrieval, orchestration, policy-based routing, tool invocation, evaluation harnesses, and lifecycle observability.
Set the standards for what "production-ready agent" means here — testability, rollback safety, cost ceilings, failure modes, human-in-the-loop boundaries — and hold the bar in review.
Take on the hardest parts of the build yourself. This is a hands-on role; you are expected to be in the code.
Build and extend the abstraction layers that let teams integrate AI capabilities cleanly and safely across our services.
Define the shared libraries, patterns, and guardrails other teams build on, and drive their adoption — a pattern nobody uses isn't a pattern.
Treat responsible use of AI on sensitive data as a hard engineering requirement, and translate privacy, security, and compliance constraints into concrete architecture rather than deferring them.
Own full-stack delivery in TypeScript/Node.js and Python: service and API layers, data-processing jobs, and the internal interfaces on top of them.
Leverage modern cloud infrastructure, event-driven patterns, CI/CD, and observability to deliver scalable AI-native systems.
Own deployment, monitoring, and troubleshooting in production, including on-call, and improve the operational posture of what you inherit.
Partner directly with product, operations, clinical operations, and business leaders as both technologist and trusted advisor — helping define which use cases are worth building and which aren't.
Lead design sessions, proofs of concept, and build-with sessions alongside the people who'll use the workflows, building trust and adoption as you go.
Communicate trade-offs, risks, and recommendations clearly to technical and non-technical audiences, up to and including the executive team.
Influence roadmap and prioritization with a clear-eyed read of technical risk, sequencing, and cost.
Own the evaluation strategy for your domain: define the metrics, test harnesses, and evaluation plans that measure agent accuracy, latency, safety, and cost-effectiveness.
Instrument the systems so their behavior is legible after the fact, not just at demo time.
Iterate rapidly on data, feedback, and changing requirements — and kill approaches that aren't working, early and visibly.
Mentor and grow engineers through code review, design review, pairing, and direct feedback; make the people around you measurably better.
Craft reusable patterns, documentation, and best practices that raise the engineering bar beyond your own team.
Anchor our internal community of practice around AI-native and agentic engineering.