AI Automation Lead, Back-Office Operations

About Sevaro :

Sevaro provides telestroke and teleneurology care to hospitals across the country. Behind every consult sits a back office: credentialing physicians at dozens of hospitals, paying them correctly every month, scheduling round-the-clock coverage, managing contracts, and invoicing our hospital clients. We are moving that work out of spreadsheets and inboxes and into systems we build and run ourselves.

The role :

You will build the AI systems that do this work, and the platforms our staff run it from every day. Each one has two halves: the automation that does the work, and the platform where a person reviews it, approves it, and answers for it. You own both, from sitting with the team that does the job today through to a live system they trust.

This is a builder role. Your measure is simple: does the payroll team, the billing team, or the credentialing team actually do their work out of what you shipped?

What you will do :

- Learn the work first. Sit with back-office staff, map exactly how the job is done today, and find where automation removes hours without adding risk.

- Build with AI agents. Direct AI coding agents to build production apps. You set the plan, review every change, and own the result. You are the engineer of record, not a prompt typist.

- Ship it yourself. Take your work all the way to production and prove it works for the people using it, not just that the tests pass.

- Protect sensitive data. Work with real pay data, billing data and patient information under HIPAA. Know what may leave which system, and never send it anywhere it should not go.

- Check everything twice. Every dollar figure recomputed from source by a second method. Every substantial change reviewed by someone, or something, that did not build it.

- Keep it running. Read the logs when something breaks, fix the root cause, and make sure the next failure is visible instead of silent.

The level of this role :

This is an AI Engineering role that it is worth saying plainly what that means here.

- One person, many production systems. You will run several live platforms at once, each carrying money, pay data or patient data. Most companies would staff each one with a small team. Here, AI agents do the typing; you supply the judgment.

- You are the engineer, the product owner and the on-call. You decide what to build with the people who use it, you build it, you ship it, and you answer for it when it breaks.

- Judgment is the job. AI will write the code fast and sound confident when it is wrong. Your value is knowing which outputs to doubt, how to prove them, and when to stop and ask a human before anything irreversible happens.

- You build the system that builds the systems. A strong, well-kept Claude Code setup is part of the job. Improving your tools makes every project better at once.

What you must be expert in :

- Claude Code, deeply. Not just using it, but configuring it : CLAUDE.md files, hooks, skills, subagents, MCP servers, permission modes and memory. You can show us your own setup and explain why each piece earns its place.

- TypeScript and Next.js. You can read, review and fix a full-stack Next.js app, including server routes, data access and sign-in.

- Building with AI. Hands-on experience shipping real systems with AI coding agents (Claude Code or similar), and calling AI models from inside a product. You can tell when AI output is wrong, and you prove it is right before calling it done.

- AWS containers and Postgres. Practical experience running apps on ECS or similar, with a managed Postgres database, schema migrations, and secrets handled properly.

- Docker and CI/CD. You can write a clean multi-stage Dockerfile and a GitHub Actions pipeline, and debug either when it fails.

- Data accuracy. Python and SQL for reconciling messy spreadsheets and records. You know that a blank is not a zero and that last month balancing does not prove next month is right.

- Security basics. Single sign-on, permission checks on the server, least-privilege access, and how not to leak data through logs or error pages.

Strong pluses :

- Healthcare, finance or operations background: payroll, billing, revenue cycle, or provider credentialing.

- HIPAA experience, including what a Business Associate Agreement (BAA) covers.

- AWS CDK or other infrastructure-as-code.

- Microsoft 365 and the Microsoft Graph API (SharePoint, Excel, Teams).

- Playwright or other browser-based end-to-end testing.

- An eye for clean, calm interface design. These are tools people use all day.

A computer science degree is not required. A track record of shipped systems that real people rely on is.