Want to know if this job is worth applying to?
United States
Remote only
Location: Remote (US)
Virtual care only works if the infrastructure behind it does. Wheel builds that infrastructure — the systems that let telehealth run at scale for the patients, clinicians, and companies depending on it. We're not bolting AI onto old workflows; we're rebuilding them from the ground up. Our mission to put great healthcare in everyone's reach means replacing manual, ad-hoc processes with intelligent, agent-powered systems that carry real operational load. We're innovating in a live, regulated healthcare environment where getting it right matters. This is the next architecture of virtual care, and we're building it now.
A staff-level engineer who sets technical direction for agentic systems and then leads the work to ship them. You've built and operated agents in production, you know where they break, and you have opinions — held loosely, argued well — about how to build them so they hold up in a regulated environment.
You operate with the scope of a domain owner, not a task owner. You take a problem that isn't yet well-formed, define it, sequence it, and lead a group of engineers to a shipped and measured outcome. Your impact shows up in other people's work as much as your own: the patterns they reuse, the design decisions they don't have to relitigate, the ambiguity you removed before it cost the team a quarter.
You treat AI two ways at once — as a product capability you build with judgment, and as a development multiplier you use fluently — and you're the person who raises the org's bar on both.
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.
First 90 days: you have a working map of our AI platform surface area, have shipped something real, and have a point of view on where the leverage is.
First 6 months: you own a domain outright, are leading a team's technical direction within it, and there's an evaluation story for the agents you've shipped.
First year: patterns you established are in use by teams you don't sit on, and engineers point to you as the reason their work got better.
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. This is a remote, cross-functional role.
Comfort with ambiguity and a bias toward shipping measurable results.
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.
Remote-first (US), fast-moving, tight loops between engineering, product, and the business. We're an AI-forward engineering org — we expect AI in the toolchain and in the product, and we measure outcomes over activity. We handle sensitive healthcare data under strict controls; privacy, security, and responsible use of AI are engineering requirements here, not afterthoughts.
Pay Range: $185,725-$264,500 USD
Final offer amounts are determined by multiple factors including, but not limited to, the scope and responsibilities of the role, the selected candidate’s work experience, education and training, the work location as well as market and business considerations.
As an employee of Wheel, you’ll enjoy our Total Rewards Program to help secure your financial future and preserve your health and well-being, including:
Medical, Dental and Vision
Ancillary: Life, Short and Long Term Disability
401K match
Flexible PTO
Parental Leave
Stock options
Additional programs and perks
Wheel is committed to equal employment opportunities for all team members. Every decision we make regarding employment is solely based on merit, competence, and performance. We are committed to building a team that represents a variety of backgrounds, perspectives, and skills.
Research shows that underrepresented groups typically apply only if they meet 100% of the criteria listed. At Wheel, we encourage women, people of color, and LGBTQ+ job seekers to apply for positions even if they don’t check every box for the role.
