The company builds AI agents that run back-office work for healthcare providers, including document intake, prior authorizations, order entry, and payer portal workflows. Customers begin with a paid pilot and typically see real work done within weeks before expanding.
The team is small, AI-native, and growing quickly. A year ago agents opened none of the merged pull requests in the codebase; last month they opened more than half. Building the system that keeps that safe is now the top engineering priority, and this role is central to it.
Responsibilities
You will not write the automations yourself. You will build the system that creates them, runs them, and keeps them working.
Browser automation is a major part of the work: hundreds of automations run against payer and provider portals the company does not control, under business rules that shift whenever a payer or customer changes them. Agents handle most of this, investigating failures, finding root causes, writing fixes, reviewing each other's code, checking fixes in production, and completing stuck work in a live browser when needed.
Build the engineering system behind that work: the runtime agents operate in, the pipeline their changes ship through, the observability that shows agents and staff what is happening, and the guardrails that stop an unchecked agent change from reaching a customer. The same system also supports document extraction, eligibility checks, outbound calling, and future products.
The work is infrastructure engineering covering distributed systems, durable execution, deployment and versioning, observability, reliability, and security. The difference is the users: agents and the people supervising them. Agents work fast and in parallel, and they will expose every ambiguity in an interface and every gap in a guardrail.
The company is hiring two engineers. This role builds the platform. The Product Engineer role owns the customer-facing tools built on top of it.
Build the runtime agents work in: isolated sessions on on-demand cloud machines, warm pools, sandboxed browsers, and narrowly scoped access to the credentials and portals each job needs.
Improve the path from a first report to a verified fix: investigation, root cause, code change, automated review, human approval, deploy, and checking results against real business outcomes in production.
Own deployment and versioning so a workflow change ships in under a minute, runs the exact code it was built with, and rolls back cleanly without someone watching.
Build observability for a highly autonomous system: metrics, alerting, dashboards as code, execution traces, and provenance linking each result to the build that produced it.
Design CLIs, APIs, and error messages for agents as first-class users, clear enough that an agent can diagnose a problem without a person translating.
Keep long-running distributed workflows correct under failure: idempotent, resumable, bounded, and safe to replay across deploys.
Build the interfaces for supervising agents, such as approval queues, code review, and live views of running work.
Run production: debug incidents, write postmortems, and fix the class of problem rather than the single instance.
Work directly with the CTO to set the engineering standards that the human and agent team follow.
Requirements
Strong professional experience building and operating production systems. What you have built and kept running matters more than a years-of-experience figure, though this usually means several years in the field.
Solid distributed systems instincts. You consider race conditions, retries, idempotency, backpressure, cancellation, and partial failure by default, not only after a bug.
You have owned the reliability of something important: been paged for it, debugged it using metrics, logs, and traces, and fixed root causes rather than symptoms.
Hands-on experience with cloud infrastructure and deployment, including CI/CD, containers, and safely shipping and rolling back changes on AWS, GCP, or Azure.
Deep proficiency in a statically typed language. TypeScript is used daily. Strong Go, Rust, or similar experience is also welcome if you can get strong in TypeScript quickly. In either case, you use the type system to make invalid states unrepresentable and avoid escape hatches such as any or unsafe casts.
You build with AI coding agents such as Claude Code, Codex, or Cursor every day and have clear views on where they fail and what the surrounding system should do about it.
Comfort across the stack. Most of the work is backend and infrastructure, and you will also build interfaces people use to supervise agents.
High standards for readable, maintainable code and little tolerance for unnecessary complexity.
Preferred: experience with a strictly typed functional style, including algebraic data types, discriminated unions, exhaustive matching, state as explicit values, and pure logic separated from side effects. Time spent in Haskell, OCaml, F#, Rust, or Scala counts.
Preferred: experience with durable execution or workflow orchestration such as Temporal.
Preferred: experience building internal developer platforms, build and release systems, or CI/CD infrastructure that other engineers relied on.
Preferred: experience running fleets of containers or VMs, including pooling, autoscaling, and graceful rollouts.
Preferred: experience with observability tools such as Prometheus, Grafana, or OpenTelemetry.
Preferred: experience building LLM agents in production, including tool design, evaluation, or guardrails.
Preferred: experience in regulated or security-sensitive domains such as healthcare or fintech, handling sensitive data.
Preferred: based in Los Angeles, or willing to move there.
Conditions
Base salary is $150,000 to $300,000, depending on experience.
Founding-team equity is offered.
Health insurance is included.
The company cannot sponsor work visas now or in the future. You must be authorized to work in the U.S. without sponsorship, and this includes F-1 OPT and STEM OPT, since both eventually require sponsorship.
Work authorization: you must already hold authorization to work in the U.S. without sponsorship.