Full-Stack Engineer — Healthcare Operations Automation

What this role is

We build automations for healthcare operations. Things people used to do in spreadsheets, portals, and email — intake, reconciliation, reporting, review, workflow hand-offs — we replace with small, focused applications that staff actually use every day. Today it's a one-person function. You would be the second.

The work has three things in common across every project:

1. Full-stack. Back-end data processing, a usable front-end, and a deployment story. Sometimes that means a Python web app packaged as a Windows .exe for a desktop user. Sometimes a containerized webapp behind SSO. Sometimes a static HTML dashboard a clinical leader opens in their browser. The right shape depends on who's using it.
2. LLM- and agent-powered. Most new builds use LLMs for extraction, classification, rule evaluation, or multi-step reasoning. You should be comfortable designing agentic workflows — tool use, structured outputs, evaluation loops, cost and latency awareness — not just calling an API and pasting the response.
3. Deployed somewhere real. Cloud (Azure is the default) or on-prem / local, depending on the customer. You own the packaging, the install, the update path, and the feedback loop when something breaks at 7am.

What is exciting about the organization and opportunity

• You'll be on the front edge of how AI is actually used in healthcare operations. LLMs, agentic workflows, and coding agents are the default tools for new builds — not a roadmap item, not a pilot, not a slide. If you want to work where applied AI is the baseline, this is that environment.
• You'll be surrounded by early adopters. The operators across the portfolio are willing to try things, give blunt feedback, and put new tools into their daily workflow within days of seeing them. That kind of feedback loop is rare and it makes you a better engineer fast.
• Internal tools can become real products. When something we build for the portfolio turns out to be good enough to serve a broader market, we have a clear path to spinning it out as a standalone, supported company. The work you do here doesn't have to end its life as an internal tool. That option exists, and it's part of why this seat is worth taking seriously.

What you'll actually do

Representative examples, not a Wishlist:

• Take a workflow that currently lives in Excel and manual portal lookups and turn it into a web app a non-technical user can run themselves. Build the ingestion, the rules, the UI, and ship it.
• Design an LLM-based extraction pipeline for a messy document type (PDFs, scans, multi-format exports) with structured outputs, confidence scoring, and an exception queue. Make it reliable enough for production.
• Stand up an agentic workflow — a chain of LLM calls and tool invocations — that automates something a human analyst used to do in 40 clicks. Write the evaluations so we know when it regresses.
• Own deployment for a specific app: container build, Azure Container Apps / App Service setup, SSO, logs, alerts. Or the equivalent on-prem: signed executable, installer, update story.
• Pick up in-flight projects and finish them. Some are 80% done and waiting for someone to close the last 20%.
• Write the spec, the README, and the handoff doc. Respond to stakeholder review notes in writing.

What you need

Most of these. Be honest about which you haven't done.

• Shipped a full-stack application used by real non-technical users. Back-end + front-end + a deployment that someone other than you maintains.
• Strong Python. Comfortable with pandas and file I/O against messy real-world data. Comfortable with at least one web framework — FastAPI, Flask, Streamlit, Django.
• Reasonable front-end. Doesn't have to be a React specialist, but should be able to build a usable interface — React / TypeScript, or a Python-first framework like Streamlit, or well-structured plain HTML/JS.
• Called an LLM API in production code — Azure OpenAI, OpenAI, Anthropic, or equivalent. Structured JSON outputs, retries, cost awareness, basic evals.
• Understands agentic patterns in practice, not in theory. Tool use, multi-step reasoning, when to use an agent vs a plain prompt, when NOT to use an LLM at all.
• Deployed something to a cloud provider (Azure preferred, AWS / GCP acceptable) with auth, logging, and a sensible CI/CD path.
• Packaged something for on-prem or desktop use — Docker, PyInstaller, MSI, signed executables. Or has the judgment to learn it quickly.
• Can work on Windows. A meaningful share of our users and build targets are Windows-first.

Things that will help you stand out

• Built or contributed to an agentic product — something using frameworks like LangGraph, OpenAI Agents SDK, Claude Agent SDK, or a hand-rolled equivalent.
• Evaluation experience. You've built offline evals for an LLM pipeline and used them to ship safely.
• Healthcare operations technology background — EHR/EMR systems, revenue cycle, eligibility, clinical documentation, practice management, claims, prior auth, patient intake, care coordination. You don't need to be an SME, but if you've worked adjacent to this space, you will ramp dramatically faster.
• On-prem deployment experience in regulated environments — hospitals, clinics, health plans. Knows why HIPAA is not a checkbox.
• Windows-side automation chops — PowerShell, scheduled tasks, Active Directory / Entra integration.

Things that do not matter

• Leetcode-style interview performance. We'll do a take-home on realistic data.
• Distributed systems at scale, large-team microservices architecture, Kubernetes orchestration. We don't have that problem yet, and when we do, the hire won't come from this JD.
• A CS degree from a specific school. Show me code and something you shipped.
• Framework monoculture. If you've only built React apps, you'll be fine — you'll also sometimes build Streamlit apps, and that's not a demotion.

How we work

• Onsite, 5 days. The job includes sitting next to the people whose workflows we're replacing. That doesn't happen over Zoom.
• Short feedback loops with actual users. We demo work-in-progress, take notes, and iterate. If a reviewer writes a document telling you your mockup is inaccurate, the first job is to read it and agree with the parts that are right.
• Local-first is often the correct answer. Cloud is an option, not a religion. A good executable on someone's laptop beats a bad SaaS tool every time.
• We use AI to build, aggressively. You should be fluent with a coding agent (Claude Code, Cursor, Copilot, or similar) as a daily tool, not a novelty.
• We don't rewrite working things for the sake of it.
• Writing matters. READMEs, specs, handoff notes. If you can't explain what you built to a non-technical stakeholder, it isn't done.

Stack you'll touch

Python 3.10+, a mix of web frameworks (Streamlit, FastAPI, Flask), pandas, Azure OpenAI / OpenAI / Anthropic APIs, Azure Document Intelligence, Azure Container Apps and App Service, Entra ID / Azure AD, Docker, PyInstaller, TypeScript / React / Node for full-stack webapps, PowerShell on Windows, Git. On-prem targets are usually Windows or Linux VMs.

New projects lean into agentic patterns and modern AI tooling. You'll have an opinion on which framework fits which problem.

Interview process

1. 30-min phone screen. One thing you shipped end-to-end. I'll ask what broke and how you knew.
2. Take-home. A realistic small task using one of our file formats. ~4 hours over a week. You keep the code.
3. Onsite, half-day. Walk me through the take-home. Two working sessions: one full-stack / data problem, one LLM / agent design problem. Lunch with the team. Questions for me at the end.
4. References. Two people who've watched you ship.

No whiteboard algorithm questions. No system-design-at-FAANG-scale. No culture-fit improv.

What this role is not

• It is not an engineering-management track. There is no team under you today. If you need to be managing people inside 18 months, this isn't the seat.
• It is not remote. If that's a dealbreaker, don't apply.
• It is not pure ML research. We use models, we don't train them.

To apply

1. Résumé or LinkedIn.
2. One paragraph on a specific thing you shipped where you were the only engineer, or one of two. What it did, who used it, what broke in production, what you changed. Bonus if it involved an LLM.
3. A link to code if any is public. Ugly is fine.

Do not send a cover letter. If points 2 and 3 aren't there, we will assume you didn't read the posting.