Provident HealthCare Management
Website:
providenthcm.com
Job details:
Agentic AI Engineer — Healthcare AI
🏥 Excellence shouldn't depend on who happened to be working that day.
That's the problem we're solving, and it's why this role exists.
Provident Healthcare Consulting is a regional healthcare operator running multiple lines of service across the United States — skilled nursing, substance use disorder treatment, and behavioral health. Our residents and patients are elderly, medically complex, in withdrawal, in crisis, or in early recovery — people who meet us at the hardest moment of their year, sometimes the hardest moment of their life.
They deserve the same standard of care on a Tuesday afternoon as they get on a Saturday night. Every time. In every building. 💯 That consistency doesn't come from asking people to try harder. It comes from building it into the system.
We're hiring an Agentic AI Engineer — Healthcare AI to help us build it. 🚀
📍 Fully remote (US) · 📋 Contract · 👤 Reports to the Chief Executive Officer
🎯 THE MISSION
Hard-wire operational excellence into every level of the organization.
Not one department. Not one workflow. Every level.
🩺 Clinical teams · 📝 Admissions · 💊 Nursing · 🍽️ Dietary · 🧹 Housekeeping · 🔧 Maintenance · 📅 Scheduling · ✅ Compliance · 🗂️ Administration
Every function in a healthcare organization touches a resident's day, and every one of them either adds to that day or takes away from it.
Right now, too much of what makes an operation excellent lives in people's heads and in their willingness to go above and beyond. That's fragile. People leave. Shifts change. Good practice at one facility doesn't automatically show up at the next one.
We want the excellent thing to also be the easy thing — the default, built into the tooling, at every level, in every building.
🔭 THE VISION
A healthcare organization where operational excellence is engineered, not improvised:
✨ The right information reaches the right person at the right moment, without anyone having to chase it
🔄 Documentation flows between systems instead of being re-keyed by hand
📡 Patterns and problems surface as signals when they emerge, not after the damage is done
🔐 Every automated action is logged, PHI-safe, and reviewable — because trust is the product
🎓 Staff at every level practice at the top of their training instead of fighting their software
The point of all of it: more time with patients, fewer hours lost to systems that fight back. ❤️
We are not doing this to reduce headcount. We are doing this because our teams are stretched, our residents deserve better, and technology should be carrying more of the load than it currently does.
⚡ WHAT YOU'LL BUILD
🔹 Multi-agent orchestration systems — state machines, checkpointing, and human-in-the-loop approval gates
🔹 Computer Use agents that autonomously navigate web-based healthcare platforms and payer systems
🔹 Hybrid agentic pipelines that extract structured data from healthcare documents using vision-language models
🔹 MCP integrations connecting agents to EMR modules and healthcare APIs
🔹 HIPAA-compliant integration pipelines between clinical and administrative systems, with field-level data merge logic
🔹 PHI-safe audit trails, minimum-necessary access controls, and Zero-Trust agent architecture
🔹 Eval frameworks and observability dashboards that keep every agent accurate and compliant — continuously
You own design, build, deployment, and evaluation end to end. 🛠️ This is production, in a HIPAA-regulated environment — not proof-of-concept.
🚫 THE THREE NON-NEGOTIABLES
Before you apply, please confirm all three:
1️⃣ Production ownership, not prototypes. You have personally designed, deployed, and operated a multi-agent system against real users and real data for six months or more. Not a hackathon build. Not a demo. Not a fork of a framework tutorial. We will ask what broke and how you found out.
2️⃣ You've shipped inside a regulated data environment. You have built and released software handling PHI, PII, or equivalently regulated data under a formal compliance regime — HIPAA, HITRUST, SOC 2 Type II, PCI-DSS, or FedRAMP. Healthcare and HIPAA experience is highly preferred.
3️⃣ 3+ years professional software engineering, Python-primary. The number matters less than what sits behind it. We're looking for someone who has owned a system end to end — architecture, deployment, and the on-call pager — not someone who has worked well-scoped tickets inside someone else's design. We can teach you healthcare. We cannot teach judgment under production failure.
🧠 WHAT YOU BRING
✅ Expert-level Python — the primary language for everything here
✅ Production experience with multi-agent orchestration frameworks (graph-based state machines, tool-calling, agent memory)
✅ LLM API integration — prompt engineering, function calling, ReAct patterns
✅ Computer Use agent development — browser automation, GUI navigation, sandboxed execution
✅ Vision-language model pipelines for document intelligence and data extraction
✅ RAG architecture — chunking, embedding, vector databases
✅ MCP server design and tool schema development
✅ HIPAA compliance engineering — PHI de-identification, audit logging, BAA evaluation, Zero-Trust design
💡 Nice to have: Behavioral health / SUD domain knowledge · HL7, FHIR, EMR/EHR APIs · AWS HIPAA-eligible services (Comprehend Medical, HealthLake, Bedrock, Textract) · LLM observability and eval tooling · adversarial red-teaming of LLM systems
📩 HOW TO APPLY
Send your resume and a short note describing the most complex agentic system you've built — the framework you used, how you handled PHI, and how you measured accuracy.
If you want your engineering work to show up in how someone's grandmother is cared for on an ordinary Tuesday, this is that job.
✉️ vmuchnik@providenthcc.com and afeinstein@providenthcc.com
Subject line: Agentic AI Engineer — [Your Name]
Equal Opportunity Employer · Fully Remote
- #HealthcareAI #AgenticAI #AIEngineer #OperationalExcellence #BehavioralHealth #SkilledNursing #HIPAA #Python #RemoteJobs #Hiring
Click on Apply to know more.
